The Economic Effectiveness of
Mandatory Engine Maintenance for
Reducing Vehicle Exhaust Emissions
Volume II
Modeling of Inspection/Maintenance Systems
January 1972
I N SUPPORT OF:
APRAC PROJECT NUMBER CAPE-13-68
FOR
COORDI NATI NG RESEARCH COUNCIL, I NC.
THI RTY ROCKEFELLER PLAZA
NEW YORK, NEW YORK 10020
AND
ENVI RONMENTAL PROTECTION AGENCY
AIR POLLUTI ON CONTROL OFFICE
5600 FISHERS LANE
ROCKVILLE, MARYLAND 20852
TRW
SYSTtMS QfiOUfi
ONE SPACE PARK • REDONOO BEACH CALIFORNIA 90?78
SCOTT RESEARCH LABORATORIES, INC.
P. O. BOX B4I*
SAN BERNARDINO. CALIFORNIA
-------
The Econom ic Effectiveness of
Mandatory Engine Maintenance for
Reducing Vehicle Exhaust Emissions
Volume II
Model ing of Inspection/Maintenance Systems
January
1972
I N SUPPORT OF:
APRAC PROJECT NUMBER CAPE-13-68
FOR
COORDI NATI NG RESEARCH COUNCIL, INC.
THI RTY ROCKEFELLER PLAZA
NEW YORK, NEW YORK 10020
AND
ENVI RONMENTAL PROTECTION AGENCY
AIR POLLUTION CONTROL OFFI CE
5600 FI SH ERS LANE
ROCKVILLE, MARYLAND 20852
APPROVED BY t.:(~tVd ~@~
RI CHARD R. KOPPANGb
PROJECT ENGINEER
~£~
NEAL A. RICHARDSON
PROJECT MANAGER
TRW
;l1~1 SCOTI RESEARCH LABORATORIES.INC.
'1!I;J};l P. O. .ox 841.
SAN ..'UIARDINO, CALII'ORNIA --
SYSTWIG -,.
,......~ ~DJjrt: D"'QW . DCnr'lAJnn &lcACH. CALIFORNIA 90278
-------
TABLE OF CONTENTS
1.0 Introducti on and Summary. . . . . . . . . . . . . . . . . . . . 1
2.0 System Definition and Assumpsions . . . . . . . . . . . . . . . 3
2.1 Engineering Design. . . . . . . . . . . . . . . . . . . . 3
2.1.1 Inspection/Maintenance Procedures. . . . . . . . . 3
2.1.2 Inspection/Maintenance Facilities. . . . . . . . . 10
2.2 Economic Factors. . . . . . . . . . . . . . . . . . . . . 13
2.3 Cons tra i nts . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4 Program Effectiveness. . . . . . . . . . . . . . . . . . . 14
3.0 Ins pecti on/Ma i ntenance System Model Development. . . . . . . . 15
3.1 Systems Data Acquisition. . . . . . . . . . . . . . . . . 15
3.2 Inspection and Maintenance Models. . . . . . . . . . . . . 17
3.2.1 Emission/Maintenance Model. . . . . . . . . . . . . 17
3.2.2 Emission Inspection Model. . . . . . . . . . . . . 21
3.2.3 Deterioration Model. . . . . . . . . . . . . . . . 29
3.2.4 Estimate of Vehicle Fraction Rejected to
Ma i ntenance . . . . . . . . . . . . . . . . . . . . 45
3.3 Operating Procedures and Labor Requirements. . . . . . . . 46
4.0 Systems Model. . . . . . . . . . . . . . . . . . . . . . . . . 52
4.1 Emissions Predictor Model. . . . . . . . . . . . . . . . . 53
4.2 Cost Estimator Model. . . . . . . . . . . . . . . . . . . 60
4.3 Operations Model. . . . . . . . . . . . . . . . . . . . . 62
4.4 Criteria for Policy Evaluation. . . . . . . . . . . . . . 64
5.0 Model Simulation Results. . . . . . . . . . . . . . . . . . . . 67
5. 1 Engine Parameter Inspection/Maintenance Procedures. . . . 75
5.2 Emission Signature Inspection/Maintenance Procedures. . . 85
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
-------
Figure
2-1
2-2
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
3-10
3-11
3-12
4-1
4-2
4-3
5-1
5-2
5-3
5-4
5-5
5-6a
5-6b
5-6c
LIST OF FIGURES
Title Page
Basic System Framework. . . . . . . . . . . . . . . . . . 4
Enforced Maintenance Program Administration. . . . . . . 12
Relationship Between Inspection and Maintenance Models. . 18
Typical Distributions of Emissions-Related Vehicle
Attri butes . . . . . . . . . . . . . . . . . . . . . . . . 20
Idle Parameter Sensitivity to Idle CO Inspection
Criteri a . . . . . . . . . . . . . . . . . . . . . . . . . 23
Idle Parameter Sensitivity to Idle HC Inspection
Cri teri a . . . . . . . . . . . . . . . . . . . .
. . . . . 24
Idle Parameter Sensitivity to Multiple Inspection
Cri teri a . . . . . . . . . . . . . . . . . . . .
. . . . . 25
Air Reactor Malfunction Sensitivity to Dilution
Correction Factor Inspection Criteria. . . . . . . . . . 27
Emission and Idle Engine Parameter Deterioration
Histories. . . . . . . . . . . . . . . . . . . . . . . . 31
Emission and Idle Engine Parameter Deterioration
Hi s tori es . . . . . . . . . . . . . . . . . . . . . . . . 32
Illustrative Emission Decay Profiles. . . . . . . . . . . 36
Comparison of Baseline Actual and Predicted Emission
Profi 1 es . . . . . . . . . . . . . . . . . . . . . . . . . 41
Operational Flow Diagram for State Inspection Lanes. . . 47
Typical State Inspection Station. . . . . . . . . . . . . 49
Systems Model Flow Diagram. . . . . . . . . . . . . . . . 54
Emission Predictor Model Flow Diagram. . . . . . . . . . 57
Cost Estimator Model Flow Diagram. . . . . . . . . . . . 63
Emission Time Histories--Engine Parameter Inspection. . . 71
Emission Time Histories--Emission Inspection
Procedure. . . . . . . . . . . . . . . . . . . . . . . . 73
Impact of Inspection Period on Program Effectiveness. . . 74
Engine Parameter Subsystem Optimization. . . . . . . . . 76
Emission Time History for Several Parameter Inspection
Procedures. . . . . . . . . . . . . . . . . . . . . . . . 79
Power Train Emission Histories--Panel A . . . . . . . . . 80
Power Train Emission Histories--Panel B . . . . . . . . . 81
Power Train Emission Histories--Panel C . . . . . . . . . 82
i i
-------
Figure
5-7
5-8
5-9a
5-9b
5-9c
LIST OF FIGURES (Cont'd)
Titl e
Page
Sensitivity of the Figure of Merit to Mode Emission
Inspection Criteria. . . . . . . . . . . . . . . . . . . . 86
Emission Time Histories for an Idle Emission and
Adjustment Program. . . . . . . . . . . . . . . . . . . . 88
Power Train Emission Histories--Panel A . . . . . . . . . . 89
Power Train Emission Histories--Panel B . . . . . . . . . . 90
Power Train Emission Histories--Panel C . . . . . . . . . . 91
i i i
-------
Table
2-1
3- 1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
4-1
5-1
5-2
5-3
5-4
LIST OF TABLES
Titl e
Summary of Emission Mode Diagnostic Sensitivity
to Parameter Malfunction or Maladjustment. . . . . . . .
Summary of TRW Data Related to Study Program. . . . . .
Summary of Parameter Survey Vehicles with CO at
30 mph .::. 2.5 Percent. . . . . . . . . . . . . . . . . .
Idle Adjustment Deterioration Rates. . . . . . . . . . .
Regression Analysis Study of HC Emission Deterioration
Factors. . . . . . . . . . . . . . . . . . . . . . . . .
Regression Analysis Study of CO Emission Deterioration
Factors. . . . . . . . . . . . . . . . . . . . . . . . .
Subsystem Proration of Maintenance Effectiveness. . . .
Comparison of Estimated Subsystem Deterioration Rates
Using Several Data Sets. . . . . . . . . . . . . . . . .
Franchised Garage Inspection/Maintenance Labor Times
and Equi pment . . . . . . . . . . . . . . . . . . . . . .
Regional Weighting Function. . . . . . . . . . . . . . .
Summary of Most Cost Effective Inspection/Maintenance
Procedures. . . . . . . . . . . . . . . . . . . . . . .
Optimum Inspection Strategies 4-Year Average Emission
Reducti ons ... . . . . . . . . . . . . . . . . . . . .
Optimal Engine Parameter Subsystem Inspection Strategies
4-Year Average Emission Reductions. . . . . . . . . . .
Optimal Emission Signature Inspection Strategies. . . .
iv
Page
9
16
28
34
38
39
42
43
51
65
69
78
84
93
-------
PREFACE
This report consists of three volumes entitled: liThe Economic
Effectiveness of Mandatory Engine Maintenance for Reducing Vehicle
Exhaust Emissions." The following are the titles given for each volume:
. Executive Summary, Volume I
. Modeling of Inspection/Maintenance Systems, Volume II
. Inspection and Maintenance Procedures Development, Volume III
The first volume summarizes the general objectives, approach and
results of the study. The second volume presents the analytical model-
ing of a mandatory inspection/maintenance system and simulation results
obtained using that system model. The experimental programs conducted
to develop input data for the model are described in Volume III.
The work presented herein is the product of a joint effort by TRW
Systems Group and its subcontractor, Scott Research Laboratories. TRW,
as the prime contractor, was responsible for overall program management,
experimental design, data management and analysis, and the economic-
effectiveness study. Scott conducted the emission instrument evaluation
and acquired and tested all of the study vehicles. Scott also provided
technical assistance in selecting emission test procedures and in
evaluating the test results.
v
-------
1.
INTRODUCTION AND SUMMARY
1.1
INTRODUCTION
The objectives of this project were to determine the effect of manda-
tory vehicle inspection and maintenance on vehicle exhaust emissions and
to identify the more cost effective procedures for diagnosing and restor-
ing to specification those maladjusted or malfunctioning engine components
which most significantly affect emissions. This volume describes the
development of the inspection/maintenance (I/M) economic effectiveness
model and the study results. The model provides a consistent and system-
atic framework for evaluating the economic effectiveness of inspection and
maintenance procedures using data developed in the experimental portion of
the study (see Volume III).
The model is sensitive to the following variables:
. The total and discounted investment costs, annual inspection
costs, system operational costs, vehicle maintenance costs,
and user inconvenience costs.
. The technological complexity introduced into the inspection/
maintenance program with emphasis on the degree of inspection
automation.
. Private, public or mixed ownership/management of inspection
and maintenance stations.
.
Inspection interval, the number, size and locations of
inspection stations; and manpower including skills and
training requirements.
. The existing voluntary maintenance program effectiveness.
The study involved the following work sequence:
. A literature survey was made and direct discussions were held
with automobile manufacturers and automobile associations
to acquire applicable cost and operational data and to identify
candidate inspection/maintenance procedures.
. Statistical evaluations of acquired data were performed to
develop models of inspection procedures, maintenance effective-
ness and deterioration and capital costs.
. An internally consistent system model was constructed using
both contract developed and acquired data to evaluate the
economic effectiveness of four candidate inspection/maintenance
strategies.
-------
1.2
SUMMARY
The following conclusions were drawn from the economic-effectiveness
study:
. The six most effective engine parameters to maintain in a
mandatory inspection/maintenance program are the three idle
adjustments (air-to-fuel ratio, rpm and timing), elements
of the ignition system when causing misfire and the induction
system components including positive crankcase ventilation
valve and air cleaner. The air injection (reactor) system
should also be inspected and maintained on vehicles equipped
with this type of pollution control equipment.
.
Inspection and maintenance of the idle adjustments was found
to be a very cost effective procedure for controlling carbon
monoxide emissions. Typical average reductions over a four
year period are between 10 to 15 percent for carbon monoxide
and 2 to 3 percent for hydrocarbons. Oxides of nitrogen
emissions increased by 4 to 7 percent.
.
Control of both hydrocarbon and carbon monoxide emissions
requires inspection and maintenance of the ignition and induc-
tion systems in addition to the idle parameters. Optimum
inspection/maintenance procedures yield a typical average
emission reduction over a four year period of 15 to 22 percent
for hydrocarbons and 20 to 33 percent for carbon monoxide.
Oxides of nitrogen emissions are increased from 3 to 5 percent
by this treatment.
.
Maximum emissions reductions are achieved with the direct
inspection and maintenance of the ignition and induction
system components in addition to the idle parameters. The
cost of this inspection is relatively high compared with the
resulting emissions reductions, making this procedure less
cost effective than the inspection and maintenance of idle
parameters.
.
State inspection lanes are almost always more cost effective
than franchised garages.
. The most cost effective inspection
frequency is once yearly.
.
Nondispersive, infrared emission measurement instruments are
preferred for state-lane emission inspections.
Cost effectiveness considerations must include the effect of
decreasing one emission species while increasing another.
For example, a substantial decrease in carbon monoxide by
leaning fuel-to-air ratio will result in a moderate increase
in oxides of nitrogen.
. Air reactor and engine modification controlled vehicles res-
ponded similarly to the inspection/maintenance procedures
evaluated.
.
2
-------
2.
SYSTEM DEFINITION AND ASSUMPTIONS
The general framework for the development of the system effectiveness
model is discussed in this section. The model focuses upon those system
elements which affect the ordinal ranking of alternate inspection/mainten-
ance procedures when their economic effectiveness is compared. These
elements are described in Figure 2-1 and are summarized below:
.
Engineering design
.
Economi c factors
.
Design constraints
. System effectiveness
The shaded areas indicated
not studied either because
considered minimal or due
on the figure denote those variables which were
their impact on selection of procedures was
to lack of data.
2.1
ENGINEERING DESIGN
As shown in Figure 2-1, engineering design involves the definition of
inspection/maintenance procedures and inspection facilities configurations.
These will be discussed below.
2. 1 . 1
Inspection/Maintenance Procedures
Optimum inspection procedures are to be defined based upon two approaches:
.
Direct diagnosis of engine parameter maladjustment or malfunc-
tion using conventional or more sophisticated garage type
equipment.
Inferences of engine maladjustments or malfunctions from the
measurement of exhaust emissions under various engine load
conditions.
.
Two limited maintenance strategies--"predeterminedll (specified) and
lIadaptivell were considered for use with these two inspection approaches.
The inspection approaches and maintenance strategies will be discussed in
further detail below.
3
-------
ENGINEERING DESIGN
ECONOMIC FACTORS CONSTRAINTS PROGRAM EFFECTIVENESS
. COSTS . ENVIRONMENT . TOTAL SYSTEM COST
CAPITAL EQUIPMENT DEMOG RAPH IC
LAND CLIMAtiC
USER INCONVENIENCE N:~~~i;:tf:{
OPERATIONS
MATERIAL
tRA*N)~~>
. TIME . TECHNOLOGY . RESULTANT EMISSION
LEVEL
USE R TIME STATE-OF-THE-ART
INSPECTION TIME ji:P~~PA$t$
MAINTENANCE TIME
. PROCEDURES
INSPECTION TYPE
PASS/FAIL CRITERIA
MAINTENANCE TYPE
INSPECTION INTERVAL
Rt.t;:~Ai3:i4ttY
. CONFIGURATION
INSPECTION FACILITY - NUMBER,
LOCATION, SIZE, DESIGN
MAINTENANCE FACILITY. NUMBER,
LOCATION
INFORMATION STORAGE/RETRIEV Al
ftEtGioNAi:; P.ART:it:~p~i:Nt:
. BENEFITS .
1!~i~!I~~~~I'I~lli
SOCIAL
METHOD OF FINANCING
USER INCONVENIENCE
<~G$t~~:ft~:N:
+:>
MO DE L
Figure 2-1. Basic System Framework
-------
Inspection Procedures
The inspection procedures to be evaluated must satisfy the following
criteria:
.
Implementation must be cost effective
Procedures must identify with high probability the malfunctions
which affect vehicles emissions
.
.
Economic impact on a single vehicle owner must be small.
The criteria imply that inspection must be used to limit the mainten-
ance options open to the repairing agency. This will protect the vehicle
owner by limiting maintenance to those repairs and adjustments shown to be
cost effective. It is also essential that the engine parameters which are
more costly to restore to specification are correctly diagnosed with a high
degree of certainty. For example, idle maladjustments may be diagnosed with
a lower level of reliability than, for example, the malfunctioning of the
pump in an air injection system because the cost impact of an idle adjust-
ment is significantly less than pump replacement.
Car owner cost and inconvenience must be minimized to encourage the
cooperation of the motoring public. Only those malfunctions shown to occur
in five percent or more of the vehicles in the field are addressed. Fre-
quently the malfunctions which occur at lower frequencies do not contribute
significantly to overall emissions and in some cases may be costly to
diagnose and repair. For example, carburetor main and power metering cir-
cuit malfunctions are not evaluated since both a dynamometer and a carbon
monoxide emission instrument are required at the repairing agency to
effect a reliable diagnosis. Even with this equipment, a lengthy inspection
is required to isolate a specific carburetor problem since the diagnosis
may be confounded by plugged air cleaners, PCV systems or leaking exhaust
valves.
In addition to cost, the vehicle owner can be significantly inconveni-
enced if his vehicle must be reinspected after mandatory maintenance has
been performed. Rather than impose this requirement, it will be assumed
that a vehicle emission surveillance program, such as described in Ref. 1,
is concurrently being conducted. Vehicles would be randomly selected from
5
-------
the general population and inspected to monitor the emission and engine
maintenance states resulting from an enforced inspection/maintenance
program. This will provide data on the effectiveness and durability of
the maintenance being performed.
Inspection procedures are to be developed for use in both franchised
garages and state inspection lanes. With franchised garages, specified
inspection procedures are applied prior to maintenance. Those vehicles
which do not comply with quantitative specifications (i.e., fail a specified
performance level) are then maintained, usually directly following the
inspection process, by the same service organization. The advantages of
this approach are:
.
No large, new capital expenditures are required to initiate
the inspection/maintenance program
.
The vehicle owner is only inconvenienced once since inspection
and maintenance are performed by the same organization.
Some control over the inspection/maintenance procedure is
because an unbiased inspection as might be performed by a
out profit motivation may not have been made.
lost in this case
state agency with-
A state-lane inspection provides the benefit of high vehicle through-
put but at the expense of less comprehensive diagnosis. Labor costs
associated with direct parameter inspection dominates inspection costs and
therefore suggest automating the state lane inspection procedure as much as
possible. The start-up and capital equipment costs of the state lane system
therefore are high.
The following ground rules were established for selecting inspection
procedures:
.
Engine parameter inspection procedures of both short duration
(approximately 3 minutes) and long duration (approximately 30
minutes) are to be defined.
.
Engine emission signature inspections are to include both loaded
(dynamometer) and unloaded (static) operating conditions.
6
-------
Engine parameter inspection procedures to satisfy the short duration re-
quirement will have to be highly automated and, therefore, are likely to be performed
in a state-lane inspection system. It does not appear realistic to use
currently available commercial equipment to perform three-minute engine
parameter inspections because of the time required to attach instruments
directly to the power train. However, direct engine parameter inspection
within a state-lane system may be feasible if remote sensing equipment is
hypothesized. A sensor which is inserted into the tail pipe and measures
idle fuel-to-air ratio, idle rpm and misfire would appear to be technolog-
ically feasible and is evaluated for use in a state-lane inspection system.
RPM, for example, might be sensed by accoustical pressure; idle fuel-to-air,
by an emissions measurement; and misfire, by a radio frequency sensor.
Idle emission signature measurements
performing a short duration inspection.
can be inferred from an idle HC emission
are also evaluated as a means for
Both timing and rpm maladjustments
measurement.
More complex engine parameter inspection procedures are required to
diagnose electrical and induction (fuel-to-air related) subsystem mal-
functions. A commercially available engine electronic analyzer is required
to inspect the primary and secondary ignition subsystems. Simple tests
which are available to diagnose some of the components affecting fuel-to-
air ratio (i.e., PCV system, air cleaners, and air injection pumps) were
evaluated in this study. No simple, reliable procedures are available for
inspecting off-idle carburetor metering malfunctions and the cost of repair-
ing these malfunctions is usually high. Therefore, these malfunctions are
not considered in this study even though the engine parameter survey,
Volume III, indicated that a moderate number of these malfunctions exist
in the vehicle population.
An alternate inspection procedure using emissions measured at various
engine loadings was also studied. Statistically designed experiments (see
Volume III) showed that high HC emissions (greater than 350 ppm) under
moderate engine loading are generally indicative of a misfiring cylinder(s)
although several other malfunctions such as severely advanced timing,
failed-air injection system and rich carburetor metering can sometimes con-
fuse the diagnosis. High CO emissions under engine load are almost always
indicative of a fuel-to-air mixture ratio anomaly. This condition may
7
-------
result from restricted flows in components such as PCV, air cleaner and
air injection pump, or from excessively rich idle adjustments. Therefore,
emission measurements can at best point only to groups of failures. For
this study, failures and maladjustments have been grouped into three
subsystems: ignition subsystem, induction subsystem and idle adjustments.
Table 2-1 shows the emission signature most likely to yield a correct
diagnosis of failures in each subsystem.
Maintenance Procedures
As a result of different inspection procedures, the maintenance state
of the vehicle will be known with different levels of precision. Mal-
functioned/maladjusted components are most precisely defined by the direct
parameter inspection. Maintenance therefore can be limited to specific
adjustments and repair. The emission signature inspection diagnosis only
groups malfunctions or maladjustments and specific components within a
suspected subsystem must be reinspected to determine the nature of the
failure and the required maintenance. The former maintenance procedure
has been defined as "predetermined" or "specified" and the latter as
"adaptive." By adaptive, it is meant that maintenance action is only taken
as a consequence of a reinspection at the component level to assess those
subsystem components which are out of specification or failed. For example,
high CO emission in a loaded mode may be the result of a component mal-
function which can be either directly or indirectly diagnosed. If failed
components are not found during reinspection the most probable explanation
for the observed high CO emissions is a failure in the carburetor fuel
metering circuits. Even though most garages cannot directly inspect for
this malfunction because a chassis dynamometer is required, the carburetor
may be repaired based on the fact that other malfunctions have not been
found which would explain the high emissions level. Where the probability
of correctly identifying a subsystem failure must be high due to high main-
tenance cost, the adaptive approach to maintenance has been selected. This
approach is similar to the key mode inspection approach proposed by Cline
in which "truth charts" are used to guide further inspection and maintenance
(~eference 2). However, the extent of maintenance required in a mandatory
inspection maintenance program will be limited on the basis of either cost
or a low probability of correct diagnosis.
8
-------
Table 2-1.
Summary of Emission Mode Diagnostic Sensitivity to Parameter
Malfunction or Maladjustment
1.0
Emission Si gnature
Subsystem/Parameter Sens i ti ve Confounding Sens i ti ve Confoundi ng
CO Mo de Parameter H C Mo de Pa rameter
* Advanced timing
I dl e, Timing I
* and misfire
Rpm I Slow idle rprn
* and misfire
I CO (i d 1 e F / A ) I Induction related
parameters; air
reactor, PCV, air cleaner Fai 1 ed vacuum
Ignition, Mi s fi re M,H
advance
Vacuum Advance M,H
*
Induction,Air Cleaner H Failed PCV, rich float H Mi s fi re
PCV M } Severely maladjusted lCD,
Leakage M rich float, air cleaner
Float Level M,H All induction related
pa rameters Advanced timing,
Ai r Reactor I ,M,H All i nducti on related I
parameters slow i d 1 e rpm
*.
Failed air reactor will also confound this diagnosis on vehicles so equipped.
I Idle.
M Moderate speed cruise load.
H Acceleration or high-speed cruise load.
-------
Predetermined maintenance is best applied where there is a high prob-
ability of correctly identifying specific malfunctions/maladjustments from
a subsystem inspection. The present study (see Volume III) has shown that
idle maladjustments or malfunctions occur frequently and are inexpensive to
repair even if the diagnosis is incorrect. However, an evaluation has been
made of both adaptive and predetermined maintenance procedures in conjunc-
tion with an idle HC or CO emissions inspection and the adaptive maintenance
procedure was found to be more cost effective.
2.1.2
Inspection/Maintenance Facilities
Design of a state-lane inspection system must include consideration of
the following factors:
.
Complexity of the inspection procedure
Number of vehicles to be inspected and their density distribu-
tion
.
. Air quality criteria of the region
. The means used to enforce the program.
The complexity of the procedure affects the inspection time and the
degree of required automation. In the case of a state-operated system, the
degree of automation and)hencejthe personnel requirements are quite signi-
ficant to costs.
Inspection station sites should be located to minimize user incon-
venience. An urbanized region, the Los Angeles Basin, was selected for
this study, thus allowing the simplifying assumption in the model that
state inspection sites are uniformly distributed. The number and configura-
tion of inspection sites will be governed by the total number of vehicles
to be inspected, the inspection time and the vehicle density. Using multiple
inspection lanes rather than single lanes tends to reduce facility cost
since land and automated equipment may be shared, but will increase travel
time for the vehicle owner since the sites will be more disperse. The
greater the vehicle density the more cost effective it is to use multiple
inspection lanes because inspection sites will of necessity be near to each
other and travel time will be a less significant economic factor than the
reduced cost of the facilities.
10
-------
Demographic and inspection station data assumed for the Los Angeles
Basin are:
.
Controlled vehicle population - 4 million
Urban area - 1250 square miles
.
. State-lane positions - 15
. Average speed to and from inspection station - 22 mph.
A franchised garage system is less flexible with regard to locating
inspection sites than is the state-lane system, however, the direct capital
costs are minimal. In this study, it is assumed that additional costs re-
quired to modify our upgrade franchised garage facilities for the purposes
of this program will be reflected as an adjusted overhead on direct labor
costs rather than as a direct capital expenditure.
For the state-operated system, some of the more significant inspection
station configuration variables such as the number of inspection lanes per
site were optimized during the study. For the franchised garage system,
the number and distribution of facilities was taken as a known input to the
study.
The ranking of inspection/maintenance procedures by cost effectiveness
is not anticipated to be sensitive to regional demographic data but is likely
to be sensitive to the weighing factors applied to the individual emission
reductions in the system figure of merit. These weighing factors can be
selected to reflect either national or regional air quality criteria.
The administration of the program requires three functional activities:
.
Inspection scheduling
.
Failure reporting
. Data recording and storage.
The functional flow of these activities is shown in Figure 2-2 and is
described below. The data acquisition system requires a two-part, prepunched
card identifying the vehicle, its owner and the month in which the vehicle
is to be inspected. This card normally would be part of the vehicle regis-
tration package. The owner presents the card at a state-lane station where
11
-------
N
VEHICLE OWNER
f NSPECT ION
5CHEOULE
INSPECTION
FAILURE REPORT
STATE
A OM I N 1ST RA T ION
Figure 2-2. Enforced Maintenance Program Administration
MAINTENANCE
COMPLIANCE
REPORT
STATISTICAL
RECOR OS
-------
the information officer inserts it into a key punch. This officer also
operates a display console which verifies the data acquisition procedure.
Concurrently, an aide inserts the sensing probe in the exhaust pipe. Due
to variations in tail pipe locations and size, this task probably cannot
be readily automated.
The analog output signal from the instrument which senses either
emissions or acoustic pressure is converted to digital output, and the
results are punched on the card. The punched card is verified against the
display console and one copy of the card is returned to the owner. For
those vehicles rejected, a failure report is provided which lists the engine
parameters to be diagnosed and/or maintained. The prescribed maintenance
is performed and verified by a franchised garage on the owner's hard copy,
which is returned to the enforcement agency.
Both the failure and compliance reports are filed within a storage and
retrieval system. Periodic comparisons are made to determine if enforce-
ment action is required.
2.2 ECONOMIC FACTORS
The benefits in reduced emissions obtained from an inspection/main-
tenance (I/M) program must be compared with its cost to determine the pro-
gram's overall economic effectiveness. The relevant costs here are the
explicit and implicit costs to implement the program. Explicit costs
include expenditures to construct facilities and to perform inspection/
maintenance operations. Implicit costs are less tangible and generally
are not expressed in monetary units. They include, for example, the time
the vehicle owner spends in inspection and maintenance related activities.
Station location and configuration design were determined by considering
both types of costs. Only explicit costs are included in the costs quoted
for the different inspection/maintenance programs.
Other benefits that may be important include the influence of mandatory
engine maintenance on vehicle safety, operating economy and major engine
repair. The cost impact of these benefits were not evaluated because data
are not available which relate these effects to the maintenance actions
required by the objective of this study (i.e., type and extent of
13
-------
maintenance). One further simplifying assumption is that the cost of
providing training for both the inspection and maintenance activities
will not be included. Both of the above simplifications are not anti-
cipated to impact significantly the ordinal ranking of the cost effective-
ness of inspection/maintenance procedures because the resulting costs are
nearly constant for the system and relatively insensitive to specific
procedures.
Capital costs to be considered are those for equipment, land and
facilities. Operating costs include utilities, labor, materials, spares,
fringe benefits and general administration. In addition, the implicit
costs associated with user inconvenience will be considered.
2.3 CONSTRAINTS
Constraints were placed on system design by the contract work state-
ment. The emission inspection procedure is to be designed to obtain
diagnostic information rather than to yield emissions measurements which
correlate well with the Federal emission test procedure. Equally important
is that procedures are to be developed to reduce HC and CO emissions.
That is, only those maintenance procedures and policies which primarily
influence HC and CO emissions will be selected for evaluation. This is
because only HC and CO emissions were considered when the emissions control
equipment was developed for the 1966-1970 vehicles studied. The effect of
selected procedures on NO emissions were to be estimated after the fact.
Additional constraints relate to the social impact of an enforced inspec-
tion program. A primary criterion for system design is to impose no
extraordinary burden on anyone vehicle owner.
2.4 PROGRAM EFFECTIVENESS
Measures of system performance include the actual emission decrements
effected, the user time lost in traveling to and from the station and wait-
ing for the inspection and maintenance to be accomplished as well as the
effectiveness of the inspection (probability of correct diagnosis).
These measures of inspection/maintenance system performance were either
used to evaluate emissions abatement performance or to establish systems cost
A figure of merit then was formulated from these elements and was optimized
by selecting values for the system design variables.
14
-------
3.
INSPECTION/MAINTENANCE SYSTEM MODEL DEVELOPMENT
The purpose of this section is to describe the development of mathe-
matical models with which to evaluate the effectiveness of mandatory
inspection/maintenance procedures. Specifically, the following major
models are described:
.
Inspection and maintenance procedure effectiveness
and maintenance deterioration
. Operating procedures and their impact on direct labor and
user-inconvenience times
. Labor and user-inconvenience operating costs; and equipment,
buildings and land to capital costs
. Measures of program effectiveness including total costs,
resultant emission reductions and an overall system figure
of merit.
These models were developed through the data acquisition and synthesis
phases described below.
3. 1
SYSTEMS DATA ACQUISITION
Data required for the development of inspection/maintenance models and
for defining system constants were derived from the following sources:
. Historical data from outside sources
.
Contract-developed data.
The available data bank is summarized in Table 3-1 and consists of
data from the Environmental Protection Agency surveillance programs,
California Air Resources Board (CARB) surveillance programs, American
Automobile Association (AAA) two-year fleet test, Scott Research Labora-
tories (SRL) New Jersey cycle development data and the TRW-developed data
from the parameter survey and statistical experiments described in Volume
III. A large portion of these data was analyzed statistically to develop
inspection/maintenance models (see Table 3-1). 1966 and 1967 vehicles with
both engine modification and air reactor exhaust emission control systems
were evaluated. Primary attention was focused on data sets containing both
control types.
15
-------
Table 3-1.
Summary of TRW Data Related to Study Program
0"1
APPLICABLE Vehicle Development Degree of Sample Modeling Application
Acquisition Goal Maintenance, % Characteri sti cs APPLI CABLE
TRW DATA RE FE RE NCE
BANK * Model Type Models
Captive Ra ndom Cycle Parameter Idle Adjust. Gen. Tune-up Siz~ Year Developed
EPA (HEW) Rental ACID* 2Q 68-69 Effectiveness of
Surveil 1 ance C1ay- 500 Idle adjustments & No Ref. 10
ton Emission Signature
Ca 1 iforni a Prfvate Surveil- " 66-69 Deterioration Yes Ref. 11
ARB lance 1000
GM Idle Private Idle ad- Effectiveness of Yes Ref. 5
justment 100 228 66-68 i dl e adj us tments &
Emission signature
AAA Fleet Company Manufact- Deterioration,
Fleet urers tune 100 100 100 66-67 Effectiveness of Yes Ref. 7
up effect- General Tuneup
iveness
EPA (HEW) Industry 100 150 68-69 No --
Certification
C1 ayton ** Key Selected Emission signa- No --
Diagnostic Mode Ma 1function~ tures
SRL-New Effectiveness, No --
Jersey Private QUi ck 25 200 Diagnostics
TRW-CRC-APRAC Pri vate 33 Emi s 11 typi ca 1 100 100 22 66-70 Emission response Yes Volume III
CAPE-13 sion failures to failures, main-
modes tenance effective-
ness
TRW-CRC APRAC Private Parameter a 0 226 66-69 Frequency and ex- Yes Volume III
Cape-13 frequency & tent of malfunction
di st ri buti ons
* General tune up includes both ignition and carburetor repair and adjustment as required to restore engine
to manufacturer's specifications.
**C1ayton's analysis of HEW surveillance dati! also available..
-------
In addition, cost data for capital equipment were obtained from emis-
sion measurement system component manufacturers, labor and materials costs
from Chilton's Manual (Reference 3) and land and facility costs from a
previous TRW report on safety compliance (Reference 4). The development of
models from these data sets is discussed in the following subsections.
3.2
INSPECTION AND MAINTENANCE MODELS
Inspection and maintenance models to be developed are of three types:
. Models relating emissions reductions to maintenance treatments
.
Models relating vehicle population attributes, such as extent
of engine parameter maladjustment, frequency of malfunction,
and emission signatures to the diagnostic (inspection) pro-
cedures
. Rates of deterioration of engine subsystems and their effect
on emissions.
3.2.1
Emission/Maintenance Model
The eMission/maintenance model predicts the change in composite emis-
sions which occur when a specified engine parameter is returned to its
nominal or manufacturer1s recommended state. This model is schematically
shown in Figure 3-1 and mathematically in Equation (3.1) below.
/J.e. =
J
n
L
i-l
3a
f
c.p.
de.
P(Pi) ~ dPi
(3.1)
where:
c.p. = policy statement rejection criterion (cut-point)
P(Pi) = probability of finding a parameter at P. + dP. with a
specified inspection procedure (from en~ine plrameter
survey Section 2.0, Volume III)
de.
--.l. =
dP.
1
response coefficient defining the change in ITI emission
per change in parameter lIill (from orthogonal experiments,
Section 3.0, Volume III).
17
-------
~
o
,
a...
A :p,~:IM~r~kJN:~r~q]:lqNt:.::
P(Pj) ERRORS OF OMISSION
-
z
o
I-
«
....J
::>
a...
o
a...
ERRORS OF COMMISSION;
CUT POI NT,
c.p.1 3eT I
1 t
I I
I I
: B.i~~m"1m!~~i :
I
I
00
UJ
l/)UJ
zU
oZUJ
a...«=
::;Jz::E
a::::UJ~
I- II>
zz E
0- m
-«.:7)
l/)::EQ)
l/)
-------
Inspection procedures involve the quantitative measurement of either
engine and control device adjustments or exhaust emissions. The vehicle
is rejected or accepted based upon whether the apjustments or emissions
fall outside specified bounds. Inspection of engine and control device
adjustments involves the steps shown in Figure 3-1 (steps A and B). The
distribution function for the deviations of engine parameters p, from manu-
1
facturers specifications expected in a general population is shown at A.
An inspection rejection level or cut point for the parameter P. is also
1
indicated. Those vehicles in which the maladjustment of parameter Pi
exceeds the cut point level would be rejected to maintenance. The effect
of adjusting parameter P. to specification upon exhaust emissions is shown
1
in B. Attributes of the population rejected in A are combined with the
emission influence coefficient of B to predict the emission reduction
achieved by this inspection/maintenance process.
Evaluating the effectiveness of inspection based upon measuring
exhaust emission involves a further step shown in Figure 3-1C. The dis-
tribution of emission levels measured using diagnostic modes or cycles for
a population of cars in shown. Again those vehicles w1th exhaust emissions
above the cut point level would be rejected to maintenance. The distribu-
tion of maladjustments of parameter P. for this rejected population is shown
1
in A. Similar distribution curves would result for other parameters which
have a significant impact on emissions. Again the effect of the inspection/
maintenance process upon emissions is obtained using the new distribution
shown in A with the influence coefficient of B.
An objective of the study will be to determine the placement of
inspection cut points which optimizes the cost-effectiveness of the process.
The upper limit used in the integration to determine emission reductions is
the 3a value of the parameter. This accounts for the 99.9 percent of the
population and probably is a reasonable limit for failures or malfunctions
in operable vehicles.
Distribution functions of the type just described have been developed
from sample vehicle populations which reflect in-use, exhaust emission con-
trolled vehicles. Some typical distributions are shown in Figure 3-2.
Data from the Parameter Survey (see Volume III) and the General Motors Idle
Adjustment Program (Reference 5) were used.
19
-------
50' CARBURETOR FLOAT LEVEL 50
40 40
30 30
20 20
Z
0
I- 10 10
-<{
...J
~
a...
0 0 0
a...
u. 50 100 150
o FLOAT LEVEl, IN x 10-2
-;R
0
N >-'
<:) IDLE HC
u 20
Z
w
~
a 16
w
IX
u.
12
8
4
0 100 400
200 300
IDLE HC EMISSIONS, PPM
Figure 3-2.
AIR CLEANER RESTRICTION
1\
'
\ y 40
'--
50 100 150
AIC RESTRICTION, DEGREES
25
20
15
10
x- 1. 26
5
0
l.q 2.0 3.0
CO AT 45 MPH, %
Typical Distributions of Emissions-Related Vehicle Attributes
-------
3.2.2 Emission Inspection Model
Figure 3-1C illustrates the distribution of specific mode or cycle
exhaust emissions for a population of vehicles. The vehicle fraction
rejected by the inspection will usually contain multiple parameter
ma 1 functi ons: l ]
P(em) = f P(Pi' P2' ..Pn)
c. p. c. p.
(3.2)
where:
P(e )
m
c.p.
P(Pn) = set
= vehicles with emissions >e rejected by cut point, c.p.
m
of parameters within P(e ) to be maintained
m
The cutpoint on the distribution represents the levels of emission at
which vehicles will be rejected. The distribution of the set of para-
meters to be maintained, P(Pn)' for the emission inspection rejected
population is shown schematically in the dashed curve on Figure 3-1A.
Some of the vehicles which were rejected by the emission inspection
procedure are in the distribution which falls to the left of the
optimum parameter cutpoint signifying that repair of these vehicles
will not be cost-effective. These errors are termed IIcommissionll errors.
The region outside of the parameter distribution P(Pn) but within P(Pi)
to the right of the cutpoint represent those vehicles which have been
permitted to pass the emission inspection but have excessive parameter
deviations from nominal. These errors are called lIomissionll errors.
The implication here is that an emission inspection does not uniquely
identify individual maladjustments but points to failures within families
of related engine parameters. These families may be classified as a sub-
system. This is consistent with the fact that the diagnostic modes were
shown to point to more than one out-of-specification parameter in the
statistically designed experiments (see Volume III). The development of
the relationship between emission inspection signatures and the subsystem
maladjustments diagnosed is presented below. Carbon monoxide emission data
were obtained from the TRW parking lot survey and hydrocarbon emission from
the General Motors data. These data were used to develop the relationships
expressed by Equation (3.2).
21
-------
Idle Adjustments
GM data were
shown in Equation
as fo 11 ows :
used to develop a functional relationship of the type
3.2 for the idle adjustments. The procedure used was
. The idle adjustment deviations from manufacturer's specifica-
tion were developed from the as-received and as-adjusted data.
.
Inspection emission mode(s) cut points were systematically
s"e 1 ected and the rej ected veh i c 1 es were sorted from the popu-
lation.
.
Statistical attributes of the idle parameters (mean, standard
deviation and failed fraction) were developed for the rejected
fractions.
The effects of applying idle mode CO and HC emission cut points are
shown respectively in Figures 3-3 and 3-4. They identify the mean devia-
tions in the rejected population of idle fuel-to-air ratio measured as
idle CO, timing and idle rpm. For example, applying a cut point of 4% in
an idle CO emission inspection (Figure 3-3) rejects vehicles with an average
idle CO deviation from the specification of about 4.2%. Similarly, an idle
HC cut point of 350 ppm (Figure 3-4) is required to reject vehicles with
approximately the same average deviation of idle CO adjustment, but this
procedure rejects only 55% as many vehicles as would have been rejected at
the ICO cut point. An idle HC emission measurement however is more effective
than the idle CO in diagnosing rpm and timing deviations. Average devia-
tions are 0.9 degree and 12 rpm greater by using an idle HC rather than an
idle CO emission cut point. For engine modification vehicles, there is a
sharp break in mean value of the idle adjustment parameters at about 300
ppm where further increases in cut point do not significantly influence
those values.
Since a 300 ppm idle HC cut point appeared optimum, a dual idle emis-
sion inspection was evaluated using this value and a range of idle CO cut
points as shown in Figure 3-5. The effect of a dual mode inspection is to
reject more vehicles at lower mean values of the parameters than idle CO or
HC emission inspections taken singularly at identical cut points. The
vehicle rejected fraction for dual emission inspections is not the sum of
the rejections calculated from the individual emission inspections since a
22
-------
5.0
N
W
#.
i4.0
o
t-
«
>
UJ
C\
03.0
u
UJ
-J
C\
2.0
~CLES R
~EJ~--
GM DATA
ENGINE MODIFICATIONS
70 #.
..
C\
UJ
t-
U
UJ
.., 1.5~
UJ
~
UJ
60 ~ ~
-J <.9
u UJ
C\
I ...
UJ Z
>
C\ 0
Z t-
«
50 <{ 1.0>
............... z UJ
o C\
" t- <.9
<{ Z
> ~
UJ
C\ --
40 ~ 0.5
~
4.0
2.5
Figure 3-3.
"-.
""'-......
3.0
3.5
Idle Parameter Sensitivity to Idle CO Inspection Criteria
IDLE CO CUT POINT
-------
GM DATA
ENGINE MODIFICATION CONTROLS
90
TIMING --
6 60
80
2.0 V')
UJ
UJ
#- e:::
RPM -- Z 0
Z 0 UJ
5 0 50
0 I-
70 ~ '
I- Z
« > Q
UJ
> 0 I- if.
UJ --- IDLE CO «
N 0 ~ 1.5
+::> 0 0.... > 0
e::: UJ UJ
U 4 0 40 I-
UJ 0 U
~ 60 w
o Z ....
w
~ e:::
V\
I- UJ
~
u
3 1.0 30 ~
VEHICLES w
'" ~EJECTED - 50 >
............
"""'-
---
250
300
350
40
0.5
20
IDLE HC CUT POINT
Figure 3-4.
Idle Parameter Sensitivity to Idle HC Inspection Criteria
-------
80
4
VEHICLES REJECTED
--
70 4.0 1.4
~ ~
Q. TOTAL VEHICLES
~
~
0 60 1.3
*
..
Q
w ~
.....
N U 0 Z
(J"J w U ~
~ 4
w
~ 50 3.0 1.2 I-
VJ
W
...J
U
:z::
w
>
40 1.1
-
30
GM DATA IDLE HC > 300 PPM ENGINE MODIFICATION CONTROLS
2.5
3.5
2.0
4.0
1.0
3.0
IDLE CO CUT POINT
Figure 3-5.
Idle Parameter Sensitivity to Multiple Inspection Criteria
-------
number of vehicles will fail both tests. The total of the vehicles re-
jected is the sum of vehicles failing each inspection less the fraction
which have failed both criteria as shown schematically in Figure 3-5.
Air Injection System
Diagnostic modes sensitive to malfunctions of the air reactor emission
control system are the closed throttle HC and CO modes and loaded CO modes
(see Volume III, Section 4). Of the closed throttle modes, the HC emissions
during deceleration should be most selective of air reactor failures as
indicated by its response relative to the other parameters evaluated.
Alternatively, Cline (Reference 6) has suggested that the dilution factor
has strong diagnostic content since this parameter reflects excess air
added to the exhaust. A dilution factor near unity as opposed to a typical
value of 1.3 would indicate a malfunctioning or failed system. There were
seven (6%) indicated cases of substantial air reactor system malfunction
within the GM data set, a cut point of 1500 ppm on the 50-20 deceleration
mode would have found only 4%. The most effective emission signature is an
idle mode inspection for dilution correction factor which rejects vehicles
at values less than or equal to 1.0 (Figure 3-6).
Because severe idle maladjustments can reduce dilution factors to near
unity on otherwise nominally performing systems, a mode substantially out
of the idle carburetor circuit might offer some improvement. The 30 mph
dilution factor mode was found to make no errors of commission using a cut
point of 1.2, although two vehicles with poor idle performance were omitted
(Figure 3-6). For the economic-effectiveness study, an idle diagnosis for
air reactor malfunction using a dilution correction factor cutpoint of
1.0 will be used for the emission inspection since it is shown to yield the
highest probability of a correct diagnosis.
Air Cleaner, PCV and Air Pump
The emission signature response coefficients (Vol. III) showed that the
more highly loaded CO modes are most sensitive to PCV and air cleaner re-
strictions. An analysis of the parameter survey data in Table 3-2 indicated
that 18% of the 136 vehicles were rejected with a cut point criteria of 2.5%
using the 30 mph CO mode. This mode appears to have fair diagnostic sensi-
tivity to PCV and air pump, identifying 39% and 60%, respectively, of those
26
-------
)
J
V)
LU
-..I
U
I
LU 18
>
a:::
o
I-
U
«
UJ
a:::
a::: 14
«
u.
o
I-
Z
UJ
U
a::: 10
UJ
a...
26
DILUTION FACTOR = 6HC+C~5+C02
22
-----
IDLE MODE
6
--------
-
2
1.0
1. 05 1. 1
DILUTION CORRECTION FACTOR USED FOR REJECTION
VEHICLES REJECTED
VEHICLES ACTUALLY FAILED
30 MPH
MODE
1.2
1.3
Figure 3-6.
Air Reactor Malfunction Sensitivity to Dilution
Correction Factor Inspection Criteria
-------
Table 3-2. Summary of Parameter Survey Vehicles with eo
at 30 mph ~ 2.5 Percent (Vehicle Data Set = 136)
I Float *
' Vehicle eo at 30 mph Ai r el eaner pev Deviation
i **
: Number (%) (% Blockage) (I n . H20) (in.) Air Pump
107 5.0 13 -0.2 0
110 5.0 0 @ NRt
124 3.2 -0.2 0.03
133 5.0 @ NR
134 ! -0.5 NR
136 -0.2 0
140 3.0 -0.4 NR
146 5.0 -0.2 @
147 3.0 3 -0.1 0
151 2.5 0 2.5 > 0 > 0.06 ~ 0.2%
I Cri teri a
* Positive values indicate rich float setting
**Change in CO in going from the disconnected to connected state
t NR, not recorded.
Note: Circled figLwes show items whi ch failed cut point criteria.
28
-------
vehicles with failures. For example, a total of 18 PCV failures were
found of which seven were identified by the CO emission inspection (about
39%). Increasing the cut point to 3.0% decreased the number of PCV and
air pump failures identified by one unit each with only a 3.5% reduction
in total rejected vehicles. It would, therefore, appear that a cut point
of 2.5% is near optimum and it is the one evaluated within the economic-
effectiveness study.
Misfire
It is assumed that significant levels of misfire (i.e., greater than
2.5%) will be diagnosed with either an idle or highly loaded mode HC emis-
sion measurement. A 2.5% misfire rate will result in a 250 ppm increase
in an otherwise nominal HC emission signature. Assuming the malfunction
is randomly distributed throughout the vehicle population, this increase
in HC emission would result in identifying at least 50% of the vehicles
which misfire at idle using an idle HC cut point of 400 ppm.
A more sophisticated inspection would involve an HC emission inspection
at load where incipient misfire at idle is exposed. Approximately twice as
many vehicle misfires would be diagnosed with this procedure. A loaded
mode HC cut point of 300 ppm would be expected to identify approximately 50%
of those vehicles which would misfire at idle. The GM data indicate that
all misfire malfunctions may be identified with cut points as high as 400
ppm. Therefore, both values will be investigated in the economic-effective-
ness study.
3.2.3 Deterioration Model
In an attempt to obtain
data sources were evaluated:
deterioration of maintenance data, several
The AAA fleet program conducted on 1966 air injection and 1967
engine modification vehicles (Reference 7)
. The CARS vehicle emissions surveillance data for model years
1966-69 (Reference 11).
.
In the former program, two fleets of company vehicles underwent
manufacturers' recommended maintenance every 12,000 miles or when the driver
complained. Approximately fifty vehicles were tested in each model year
every 4000 miles. The AAA allowed TRW direct access to its basic data in
29
-------
order to obtain mode emissions and idle parameter adjustments to supple-
ment the reported data of Reference 7.
The CARB surveillance data by model year were also utilized to develop
deterioration models. This program is used by California to determine
vehicle compliance with emission standards over the first 50,000 miles of
the vehicle's life. Over 1000 randomly acquired vehicles have been tested
for hot cycle emissions in this program. However, statistical data are not
available on the vehicles' general engine maintenance states.
Subsystem deterioration models were developed from the AAA data set
through statistical analyses. The sum of the effects of subsystem deterior-
ation rates was then constrained to fit the ARB surveillance data.
AAA Fleet Data Evaluation
To assess the stability and effect of idle adjustments on emission
levels, the AAA data were plotted against cumulative mileage for a random
selection of vehicles (make/model/years). Figures 3-7 and 3-8 depict these
results. Exhaust emissions as determined using the full Federal procedure
(post-tune emissions measurements were made in the hot state). Idle rpm,
A/F and timing are shown plotted against cumulated mileage. Ordinates have
been selected for the idle parameters such that their slopes are inversely
related to emission level. To aid interpretation, the idle
adjustments are coded to the emission they most significantly effect (i.e.,
in the figures both timing and HC are solid lines). In a large number of
instances (approximately 60%) both the trends and variations in the emission
are explained in part by the idle adjustments. It is also of interest to
note that typically both HC and CO follow similar trends. This is based on
observation of the total data set, not only Figures 3-7 and 3-8. In a few
instances, unexpected emission changes occurred which opposed the trend of
the idle adjustments and which were not corrected by maintenance (see the
post-tune HC in Figure 3-8).
profiles
The following conclusions are indicated:
.
Idle rpm trends to be relatively unstable while timing appears
to be relatively stable.
.
Each parameter tends to degrade with a characteristic slope
over long periods of time (8000-12,000 miles) after an adjust-
ment.
30
-------
RPM IDLE A/F HC EMISSIONS
.... '-" '-" '" w .... .... '-"
'-" 0 '-" :;;: '-" 0 0 '-" 0
o 0 0 '" w 0 0 0 0 0
o
---
"'T1
.......
<.Q ....
c:
)
ro
w
I
'-J 00
I
CJ I'Tl I
ro :3
c-t- ....... '"
ro l/)
) l/)
-.J. --.I..
0 0
) ~
AJ
c-t-AJ -"11JJ
W .......~ r- O-m
--' 0 0.. m ,-r-"
~ }> m-tr-
........ G) »m»
:r: 0.. m", .OJJ C')
....... --' 0 <-"m
l/) ro 5i~»
c-t- ~CI)Jj ~"
0 I'Tl
) ~ , }>O
....... <.Q '" m "",
.... Z- ~o
ro ....... -t m",
l/) ~ -1}>
ro -1Z
""U ~n
'" -1:r:
AJ 00 om
) ~O n
AJ 0
:3 -}> ,}>
ro -1 "
c-t- ~ "
ro w
'" Z
) G)
W
0- I J,.. 0 ....
00 ~ w ....
tn 0 '-" 0
TIMING DEVIATION, DEG CO EMISSIONS
-------
RPM
IDLE A/F
'"
o
o
'"
ll>
o
"
ll>
o
"
o
o
'"
w
o
-<
3:
z
G)
., -""
<..0
C
.....,
CD
W OJ
I
00
0 rrl
CD 3 '"
c-+ -I,
CDU1
....., U1
00
.....,::1 3:'"
CU
w c-+cu r
N -1,::1 m
0 0.. »
::I G)
........ m",
:r: 0.. 0
-I. --'
U1CD
c-+
Orrl
....., ::I
-I, <..0 '"
CD -I. -""
U1::1 ,
CD
-u ,
cu
....., '"
cu OJ
3 /
CD /
c-+
CD /
....., w ;;tJ
'" "
3:
w
'" -""
J,.. 0
TIMING DEVIATION, DEG
'"
o
o
7 MODE HC EMISSIONS
w
o
o
-""
o
o
+
~G)
mZ
»::;
~O
cz
V>
-< -<
~~
Zm
-<
I
I
1..-
--
»
~~
~~
»
-<
»
u..
o
'"
'"
u,
w
o
7 MODE CO EMISSIONS
-------
The AAA fleet data were then placed on punched cards and the data were
segregated by the subsystem maintained (i.e., idle and idle plus ignition).
The following analyses were performed to characterize the statistical pro-
perties of both emissions and engine parameters:
. All vehicles undergoing periodic idle adjustment plus ignition
system maintenance were pooled and classified by mileage range.
Statistics on mean engine parameter premaintenance values were
developed.
The premaintenance emissions (He and CO) for idle plus ignition
maintenance were regressed against mileage and the idle para-
meters.
.
Engine Parameter Stability
Two statistical characteristics which are of interest in characteriz-
ing idle adjustment deterioration are the changes in mean values and dis-
tribution (variance) as a function of accumulated mileage over a maintenance
interval. These results (Table 3-3) suggest fairly rapid trends in the
deterioration of the idle adjustments from specified manufacturers. settings.
The data sets are admittedly small but they substantiate the trends previously
indicated in the parameter survey (Volume III). These trends are:
. Mean timing tends to become more advanced with mileage, increas-
ing between 0.6 to 1.4 degrees in 12,000 miles.
.
Idle rpm is very unstable as indicated by its large standard
deviation (~100 rpm).
Idle CO increases with mileage, mean increases of 0.6 to 0.7%
being typical in 12,000 miles.
These observations are consistent with the trend of the parameter sur-
vey data and the estimated frequency of voluntary maintenance (12,000 to
14,000 miles). The largest descrepancy is for the idle CO settings where
mean values for this parameter in the survey were found to be larger by some 200%.
This suggests that of all of the field adjustments, idle FIA is the least
accurately set. Since most garages are not equipped nor motivated to accur-
ately set idle FIA, this result is not surprising.
.
AAA data suggests that to a first approximation emission degradation
models can be approximated by a long-term, moderately shallow trend line
component due to deterioration and deposit build-up and an unstable trend
line component due to the less stable idle adjustments.
33
-------
Table 3-3.
Idle Adjustment Deterioration Rates
AM Fl eet Test
Parameter Mean Value and Distribution
----- ..... .- --- ~.. -
0- 3000 miles 9000-16500 miles
I d 1 e , CO, P 0 0.61%
Gp 0.94 2.04%
Timing, liP 0 1.45 deg
GlIP 0 3.51 deg
I d 1 e rp m, II P 0 -1 rpm
GlIP 0 97 rpm
P = mean level of idle parameter or its deviation from specification
G = estimate of the standard error of P or liP
34
-------
Therefore, a deterioration model, cpmposed of long-term degradation
which is functionally related to accumulated mileage and a more random
factor related to idle adjustment stability was developed.
n
FHC = Ko + Km Mileage + ~
j=l
K.L'lP.
J J
---
--
~
--
rings, valves, ignition,
carburetion
idle
adjustment
where:
FHC = Federal hot cycle emission
Km = Regression coefficient on mileage
Kj = Regression coefficient on the "j" idle adjustment
Pj = Deviation from specification of "j" idle adjustment
For a vehicle which has undergone periodic, tuneup maintenance, the
superimposed elements resemble those of Figure 3-9. The long-term deterior-
ation elements are represented by the monotonically increasing dashed line.
The emissions, due to the more random idle adjustment parameters, are
schematically represented as periodic noise whose rms value tends to in-
crease over a maintenance interval. At the end of the interval, the idle
adjustments are restored to manufacturers. specification and, again, they
degrade from their null points. To test this hypothetical model, a series
of regression analyses were performed with the AAA data partitioned into
air reactor (AIR) and Engine Modification (EM) populations. Regressions
performed were:
. Mileage only
. Mileage plus idle adjustments exclusive of idle CO
. Mileage plus all the idle adjustments.
If the hypothesized model is correct, the following trends can be
expected to develop:
. The regression (slope) coefficient on mileage should decrease
as the idle adjustment variables are added to the regression
analysis since changes in some of these variables result in
skewed emissions changes.
35
-------
EMISSIONS
W
0"\
...
-.-. LONG TERM
----- RANDO DETERIORATION
M IDLE AD
MEAN JUSTMENTS
EMISSION LEVEL
"
I ,
"...,
,
, ,
,,' , \ '
...",.--.~.~. -......
-.- .~, ,..'
MAINTENANCE
INTERV AL
"
,"'- I
I' '
,. ...
. '
.~. ..,..-.........-.
'.' ' "
.
...
MILEAGE
Fi gure 3-9.
111 us trati Ve
Emission 0
ecay Profiles
-.-.--.-.-.
-------
. Those parameters which are strongly correlated to
emissions should have a statistically significant
"t" statistic which is a measure of the regression
curve fit (greater than 2.0).
The following significant trends were noted:
Idle CO is always a highly significant variable and
mileage is also significant in most cases for HC
emissions.
. The remaining idle adjustments, timing and idle
speed (not shown here), are only infrequently found
to be significant variables. Significant idle
adjustment effects occur in a more or less random
manner within the total set of regressions.
. The multiple correlation coefficients and "t"
statistic values always increased when idle adjust-
ment variables were considered in addition to mile-
age when explaining emission degradation. The
regression slope coefficient on mileage decreased in
this case.
.
The following conclusions are based on these observations and the
data summarized in Tables 3-4 and 3-5.
. There is a basic deterioration rate, Km, (1.44-2.2 ppm HC
and 0-0.03% CO per 1000 miles) which is independent of the
idle adjustments (i.e., determined after setting ~Pj to
specification values).
. HC and NO emissions from air reactor controlled vehicles
were more sensitive to variations in the parameters
studied than were CO emissions. The reverse was the
case for engine modification controlled vehicles.
. The high value of the "t" statistic for the idle CO
effect upon emission degradation and the corresponding
significant decrease in the mileage regression co-
efficient indicate that this maladjustment is
significant and explains part of (20-30%) the
emission degradation originally attributed to mileage.
The coefficients determined from the larger data set with the larger
accumulated mileage range were selected to guide the determination of
deterioration profiles at the subsystem level (see shaded coefficients
in Tables 3-4 and 3-5).
37
-------
Table 3-4. Regression Analysis Study of HC Emission Deterioration Factors
(AAA AS-RECEIVED DATA SET)
:..v
::0
HC/103 MILES HC/oTiM HC/%CO -
EMISSION CONTROL N M R2
TYPE KM tM KTIMING tTIMING KICO tlCO MILES
HC AIR g}f/, \ 3.4 -0.62 2.0 22 5.6 122 21, 800 .57
HC AIR 2.4 2.1 0 0 41.8 5.2 63 13,700 .66
HC EM :J:d1 3.3 2.22 1.3 16.0 6.8 146 17,900 .60
HC EM 0.40 0.47 1.55 0.85 14.4 4.6 82 11, 300 .52
AIR - AIR INJECTION SYSTEM
EM - ENGINE MODIFICATION
tJ - THE ,"t'l STATISTIC ON THE REGRESSION COEFFICIENT, KJ (SIGNIFICANT FOR t ~ 2.0)
HC = KO + KM M + KTIMING .6PTIMING + KICO PICO + KRPM 6 RPM
WHERE HC - FEDERAL HOT COMPOSITE EMISSIONS M - THOUSANDS OF MILES
K J - REGRESSION COEFFICIENT ON VARIABLE J" P J = VALUE OF THE VARIABLE
R2 - MULTIPLE CORRELATION COEFFICIENT
N - SAMPLE SIZE
-------
Table 3-5. Regression Analysis Study of CO Emission Deterioration Factors
(AAA AS-RE CE IVE D DATA SET)
w
~
CO/103MILES CO/oT 1M CO/%CO - -
EMISSION CONTROL ICO N M R2
TYPE KM tM KTIMING tTiMING '
-------
CARB Surveillance Data Evaluation
The California ARB vehicle emissions surveillance data describe
composite, hot cycle HC and CO emission degradation with accumulated
mileage for the vehicle model years 1966 through 1970. This is the only
large definitive data set which indicates long-term deterioration rates
and, therefore, it was selected to represent the performance of a baseline
fleet subjected to voluntary maintenance. These data were used to:
. Estimate the effectiveness of voluntary maintenance as it
is currently occurring.
. Verify idle adjustment deterioration rates derived from
the AAA data.
. Characterize the performance of the baseline fleet in the
economic-effectiveness study.
The deterioration rate data from the AAA fleet program and the
estimated frequency and effectiveness of voluntary maintenance were com-
bined and used to reconstruct the ARB surveillance data analytically.
The results of the evaluation are shown in Figure 3-10 which compares
the predicted emission profile against the ARB profile. It should be
noted that the initial 12,000 miles of the ARB profile, to the extent
that it is free of major induction, engine seasoning and ignition system
deterioration effects, are indicative of emission changes produced by
deterioration of the idle adjustments. The problem was to select the
frequency and effectiveness of voluntary maintenance which would yield
the best prediction of the ARB degradation data. The reliability level
of the frequency data led to the selection of the effectiveness of
voluntary maintenance a& the unknown to be determined. A fit with the
ARB emission profiles was achieved by varying the reduction in emissions
obtained with voluntary maintenance.
Data Synthesis
A reasonable range of maintenance effectiveness and its apportion-
ment along the three subsystems was first estimated using data available
from the literature (GM idle program, Reference 5) and the results of
pre- and post-tuneup data from the statistically designed experiments of
this study (see Volume III). Maintenance effectiveness in percent
40
-------
~ 600
a..
a..
...
U
I
2.0
~
0
...
0
u
1.0
900
,
- - - - PREDICTED EMISSIONS USING
ECONOMIC EFFECTIVENESS MODEL
I I I
CALIFORNIA AIR RESOURCES BOARD
SURVEILLANCE DATA
---- - - ---
--- ----
-- ----
300
3.0
----- ----- -----.
---- -----
---
o
16 24 32
MILEAGE IN THOUSANDS
40
48
8
Figure 3-10. Comparison of Base Line Actual and
Predicted Emission Profiles
41
-------
reduction is given for emissions and ordered in Table 3-6 by an increase in
the extent of the maintenance.
A proration of maintenance effectiveness over the major subsystems
(idle, induction and ignition) can be made by subtracting the GM idle
adjustment effect from the effect of the Scott Research Laboratories'
major tuneup on 22 vehicles (Table 3-6). Making the assumption that the
residual difference is primarily attributable to maintaining the secondary
ignition system and the induction system for HC and CO, respectively, and
subjectively reflecting the engine parameter survey, the following pro-
ration was derived.
Table 3-6.
Subsystem Proration of Maintenance Effectiveness
Emission Subsystem Emission Reductions, % SRL Tuneup*
Reduction Idle Adjustment** I Ignition Induction Emissions
Reduction,%
I
I
I 12t
HC 6 7 25
CO 15 - 7 22
I NO f ot - -5 -5
I
*Based on pre- and post-maintenance of 22 orthogonal test vehicles (Volume III).
**From GM Idle Adjustment Program.
tBased on engine parameter emission response surfaces from the orthogonal tests.
The data show that significant HC emission reductions will have to
come from maintaining the secondary ignition system, whereas significant
CO reduction can be achieved by the simpler idle fue1-to-air ratio
adjustment.
To a first approximation, the subsystem deterioration rates should
be proportional to the subsystem maintenance effectiveness (i.e., a sub-
system requires maintenance in direct proportion to the degree it deteri-
orates). The emission deterioration rate data from the AAA fleet evalua-
tion and ARB emission deterioration profile adjusted for proportionality
with subsystem maintenance effectiveness are shown in Table 3-7.
42
-------
..j:::"
w
Table 3-7.
Comparison of Estimated Subsystem Deterioration
Rates Using Several Data Sets
DETERIORATION RATE, dejdM
EMISSION SUBSYSTEM AAA ADJUSTED ARB TRW STUDY
DETERIORATION
AIR EM AIR EM AIR EM
IDLE ADJUST .50 .80 0.6 0.6 0.8 1.0
HC, PPM INDUCTION .50 .50 0.6 0.6 0.6 0.7
103 MILES
IGNITION 1. 70 .90 1.5 1.0 2.0 2.0
CO, % IDLE ADJUST .008 .013 .009 .009 .016 .020
103 MILES INDUCTION 0 .030 .004 .004 .004 .006
NO, PPM IDLE TIMING 2.8 2.8
103 MILES INDUCTION -1. 6 -2.4
-------
The comparison between these two data sets is surprisingly good
considering the quality of the data and the nature of the required
approximations. The values selected for this study are shown in the
final column. They reflect a compromise between the AAA, ARB and TRW
parameter survey data.
The NO deterioration rates in Table 3-7 were estimated from the
combined engine parameter survey, orthogonal experiment and air cleaner
experiment data, presented in Volume III. It is assumed NO is pre-
dominately influenced by basic timing and loaded mode fuel-to-air ratio.
This assumption is strongly supported by the emission mode response to
these parameters derived from statistical experiments described in
Vo 1 ume II I.
Therefore, the air cleaner restriction experiments, which are free
from deviate timing influences, were used to derive a linear relation-
ship between the increased composite CO emissions resulting from induc-
tion subsystem malfunction and the associated change in NO emissions.
The resultant relationship derived from the 11 basic power trains tested
in the air cleaner experiment is:
dNO - dNO dCO - PPM
dM - - dCO dM - - 1.6 103 mil es
where:
dCO/dM = change in CO emissions per thousand miles (.004) (Table 3-7)
dNO/dCO = 400 ppm/% (weighted average of 11 power trains from
regression analyses)
The influence of timing is estimated using the rate of deterioration
of timing as indicated by the AAA data and the composite emission
response derived from the definitive orthogonal experiment (Volume III)
dNO = dNO dt = 2 8 PPM
dM dt dM . 103 mil es
where:
dNO/dt = average composite NO emission response to timing
deviation, 34 ppm/deg (from orthogonal data, Vol. I)
dt/dM = deterioration of basic timing, 8.2xlO-2 deg/103 miles
(from Table 3-3).
44
-------
The projected deterioration rates were then evaluated within the
model framework to determine the compatibility of the total input data
set. This data set consists of the following for both engine modifica-
tion and air reaction emission control systems.
. Subsystem deterioration rates
. Effectiveness of subsystem voluntary maintenance
. Response coefficients from the definitive~ orthogonal experiment
. Frequency and extent of engine parameter maladjustment and
malfunction
.
Frequency of voluntary and enforced maintenance.
The results of the preliminary studies with these data indicated
that the estimated deterioration rates for the idle adjustments and per-
cent misfire were too low. These deterioration rates were therefore
adjusted upward to those values shown in the last column of Table 3-7.
These values were used in all of the procedure evaluation studies re-
ported in Section 5.
3.2.4 Estimate of Vehicle Fraction Rejected to Maintenance
When multiple failures exist in a vehicle and are to be detected
using several inspection procedures~ the combined rejection criteria
(i.e. ~ idle CO and HC) will not fail the additive sum of the rejected
population of the two independent cutpoints since some vehicles will
have failed both tests.
A statistical analysis of the parameter survey data (Volume III)
indicated that to a good approximation most failures are statistically
independent (exceptions being the failure pairs timing-dwell and idle
rpm-idle fuel-to-air ratio both of which are weakly correlated). Apply-
ing the assumption of mutual independence allows the construction of a
relatively simple statistical model for calculating the number of vehicles
rejected to maintenance (i .e. ~ the probability of finding "X" failures
in a specific vehicle), Reference 8.
45
-------
3.3 OPERATING PROCEDURES AND LABOR REQUIREMENTS
Operating costs associated with inspection/maintenance procedures
were developed from existing data or from work structure breakdowns.
These procedures serve as the specification against which labor grade
rates are selected. These data were developed from industry standards
such as flat rate labor and parts manuals for those inspection/maintenance
procedures applied in garages. When work element times were not avail-
able, they were estimated; in some cases, these estimates were verified
from observations of diagnostic center operations. Specifically, the
following work breakdown structures were developed:
.
State-lane inspection procedures using remote sensing
instruments
Garage diagnostic and maintenance operations.
.
State-Lane Inspection
A typical operating sequence for a state-lane inspection using an
idle emission signature approach is shown in Figure 3-11. The vehicle
owner is assumed to stay with and operate his own vehicle throughout the
inspection. Two parallel functions are required: 1) data acquisition
and recording, and 2) exhaust sampling probe or acoustic sensor position-
ing. The data acquisition is part of the more general function of program
enforcement which would probably be integrated into an existing, analogous
function, such as the state licensing of automobiles. Costs may therefore
be shared between this program and the inspection enforcement program.
With a highly automated, digital information system, administration costs
would be minimal and functionally similar to those of existing, large
scale billing operations.
Therefore, a fixed direct operating charge of $.50/vehicle is assessed
for the enforcement function. The inspection procedure is estimated to
require one minute and two employees. A half-minute of slack time is
available for preparing the probe for the next vehicle. Sample line
purging could be automatically or manually performed at this time.
46
-------
OWNER
IDLE INSPECTION TIMELINES AND COST
1ST ATE EMPLOYEE
I 0.5MIN
0.5 MIN
..j::;:,.
'J
DRIVE VEHICLE I ENGINE CONDITIONING
TO AREA AND DRIVE VEHICLE
SUBMIT OWNER I AND IDLE; RECORD TEST - OUT 0 F AREA
REGISTRATION I OWNER INFORMATION
I
I
I DATA ACQUISITION
SYSTEM
I .
I 0.5 MIN 0.5 MIN
I INSTALL PROBE, PURGE LI NES
I TAKE SAMPLE WITH N2 GAS
I AND DISCONNECT
I
FIXED EQUIPMENT COSTS
OPERATIONAL COSTS
NDIR
$ 2, 000
ADMINISTRATION 1 MIN X $5/HR
DATA ACQUISITION
15, 000
MECH. NO.1
MECH. NO.2
1 MIN X $3/HR
1 MIN X $4/HR
$ 0.08
0.05
MISCELLANEOUS
TOOLS, WORK BENCHES,
OFFICE EQUIPMENT 1,000
DATA MANAGEMENT
0.07
0.50
$ 0.70
TOTAL
$ 18, 000
Fi gure 3- 11 .
Operational Flow Diagram for State Inspection Lanes
-------
The use of a dynamometer for acquiring data under engine load would
require an additional minute for setting up the dynamometer and making
a loaded mode measurement. As long as remote sensing equipment is used
(i.e., that doesn't require direct attachment to the vehicle) the above
described procedure should be insensitive to the use of either an emis-
sion measurement or an engine parameter measurement. Typical operational
and equipment costs are $.70 per vehicle and $18,000, respectively.
The labor grade of the information officer assigned to these tasks
is estimated to be slightly higher than that of an average mechanic
(approximately $4/hr.) and slightly lower for the aide (approximately $3/
hr.). Therefore, the average wage used to cost direct labor is $3.50/hr.
A burden factor of 50% is applied to cover the cost of fringe benefits
and indirect operating costs. This burden rate does not include the
amortization of capital equipment which is carried as a separate account.
Costs per inspection lane range from approximately $18,000 for a
simple idle emission inspection (Figure 3-11) to $25,000 for inspections
requiring simple dynamometer loading. The $7,000 costs include the
dynamometer plus additional data handling equipment required for the
installation.
Facilities costs are based on both land acquisition and structure.
Typically, basic site layout is priced for the following design elements:
. Land occupied by structure and employee parking (6,600 ft2)
. Land required for egress and exit from the inspection lane
(600 to 1,000 ft2) plus land required to contain the to-be-
inspected vehicles (estimated from vehicle lane throughput)
All required land is priced at $2/ft2, which is equivalent to the cost of
commercial property located in areas of high traffic flow.
A typical layout of an inspection facility is schematically shown
in Figure 3-12. Provision is made for securing the instrument and infor-
mation system when they are not in service. An environmentally control-
led, enclosed module is also provided for stores, records, on-site instru-
ment servicing, and general administration. The vehicle inspection site
is covered and elevated to keep it free of water or snow and to minimize
the discomforture of the test personnel. Estimated cost for this
48
-------
+::>
~
'""""""".,
.- .
.-
FOLDING
LOCKABLE
SCREEN
ADMINSTRATIVE OFFICE
Figure 3-12. Typical State Inspection Station
-------
structure and its services is $10/ft2. Approximately 1200 square feet
of covered area are required. Total facility costs, equipment, land and
structure, will nominally range between $44,000 to $52,000 per lane.
These investment costs, when amortized, constitute a very small fraction
of the cost per vehicle inspected.
Franchised Garage Inspection
The cost and direct labor of inspection and the associated main-
tenance in a franchised garage were established from three sources
depending on whether conventional or advanced equipment is to be used.
These were:
. Chilton's Labor Guide and Parts Manual (Reference 3)
. Inspection labor times as measured on the as-received
vehicles in the orthogonal tests (Volume III)
. Coarse operations analyses for hypothesized advanced
equipment such as remote sensing devices.
Inspection labor times for the individual parameters were measured
during the initial inspection of the orthogonal test vehicles. These
inspection times were subtracted from the total estimated job times
abstracted from the flat rate manual to differentiate between the costs
associated with inspection (instrument hookup) and actual maintenance
(adjustment or replacement). This differentiation is important in
allocating costs between 100% inspection and fractional maintenance when
studying procedures using franchised garages. All labor costs for garage
inspection/maintenance are charged at a rate of $lO/hr. This is a burdened
rate which includes overhead factors and profit. An additional $.50/
vehicle is charged f0r program enforcement (i.e., the information system
required for recording, processing, storing and disseminating inspection/
maintenance data). Labor times and equipment requirements used in this
study are summarized in Table 3-8.
50
-------
Table 3-8.
Franchised Garage Inspection/Maintenance
Labor Times and Equipment
Subsystem Component Equipment Time (hours)
Inspection Maintenance
I dl e Adj us tmen ts I CO/ rpm NDIR CO and 0.10 0.20
Tachometer
Timing Timing Light 0.05 0.18
Secondary Ignition Plugs, distribu- Scope and 0.25 1. 30
tor and wi re Dynamometer
harness
Induction Air Cleaner Pressure drop 0.05 0.2
across element
PCV Crankcase pressure 0.05 0.3
at idle
Air Injection Air flow at idle 0.15 0.6
or NDIR CO
and C02
51
-------
4.
SYSTEMS MODEL
Automotive emission control through a program of inspection and main-
tenance can be accomplished by employing anyone of a number of basic
strategy alternatives. The application of a system analysis approach,
the keystone of which is a mathematical simulation model, provides the
mechanism for thoroughly evaluating the various alternatives available.
It will be seen that such a model can be an effective management tool for
examining the impact of various policy statements on emission levels. By
describing and relating the various physical and economic components of
the system, the model serves as a device for translating specific emis-
sion policy statements into estimated emission levels and costs. The
impacts of alternative policies are thus converted into comparable
quantitative measures. By defining an objective function, the net effects
of each strategy alternative can be immediately compared with others to
assess their relative attractiveness.
For all practical purposes, the model can be viewed as an abstract
extension of a real world vehicle inspection maintenance system. The
engineering elements define the relationships between program policy
decisions and resulting emission levels. This is done through a series of
transformations linking specific pass/fail criteria on the inspected
attribute to appropriate maintenance and subsequently to resultant emission
levels. The emission data are then interfaced with the corresponding
economic information in order to evaluate system cost effectiveness. The
other salient engineering aspect of the program design is the system con-
figuration. The number, size and location of both inspection and mainten-
ance facilities required to operate the program are determined using ve-
hicle population, inspection interval and operational times. The economic
factors considered in the model include both capital and direct operating
costs. In addition to these explicit costs, consideration is given to the
costs assigned to user inconvenience. These implicit costs must be
accounted for in terms of overall system performance. Lastly, the model
assesses the impact of regional air quality requirements and present state-
of-the-art instrumentation on program goals and system configuration.
52
-------
To obtain more insight into the relationships between the various
program components, consider the system diagram in Figure 4-1; here, the
engineering and economic factors discussed above have been described us-
ing two mathematical submode1s. The emissions predictor model, shown on
the left, incorporates all of those elements involved in transforming
policy decisions to emission predictions.
The cost estimator model determines the various costs associated
with operat~ng the program. The communication between the emissions
predictor and cost estimator models is provided by the operational analysis
model. This model determines the user times involved in interfacing with
the program. Once both emission levels and costs have been obtained for
any given policy set, the data are summarized using a single performance
value, a systems figure of merit. The other important factor needed in
measuring system performance are the baseline emission profiles which
reflect the current "state of nature" (i.e., voluntary maintenance).
These then are the basic models required in translating specific control
strategies into actual emission levels which can be tested in terms of
their overall cost effectiveness. Presented in the next two sections are
detailed technical discussions of the emissions predictor model, cost
estimator model and operational analysis model.
4. 1
EMISSIONS PREDICTOR MODEL
The effectiveness of engine parameter inspection procedures depends
upon their ability to identify specific engine parameter malfunctions.
This is particularly true for the case of the emission signature inspection
where several engine malfunction interactions can interfere in identifying
specific ones. In order to provide a consistent basis for comparison,
emission reductions were calculated using basic transformations relating
mode emission levels, engine parameter settings and mean emission levels.
Mathematically. this relationship can be expressed by:
T
~e -L
m
T2
~P.~ ~e.
1 J
(4.1)
53
-------
U1
.j::::.
INITIAL
CON DITIONS
AND
BASIC DATA
, r---------,
EMISSIONS PREDICTOR MODEL I I
I COST ESTIMATOR MODEL
. REJECTION CRITERIA f--j0PERATIONS MODEL~
. AVERAGE EMISSION LEVELS I. FACILITIES: . CAPITAL COSTS
. DECREMENT AND DECAY I.QUEUING I . OPERATING COSTS
. INTEGRATED AVERAGES I. TRAVEL TIME L . USER TIME COSTS
~ :
L______---_J
BASE
EMISSION
CASE
EB
ET FIGURE OF MERIT $
~ $
EW. liE.
1 1
Figure 4-1.
Systems Model Flow Diagram
-------
where:
e = mth mode emission (e.g., idle CO)
m
~P, = ith engine parameter (e.g., idle
1
adjustment)
~e. = mean composite emission level (e.g., carbon monoxide)
J
T = transformation, mathematical relationship between parameters
(see Figures 3-1 through 3-4)
For the engine parameter diagnostic inspection, only the last trans-
formation, T2' is required in translating engine parameter inspections to
mean emission levels. The emission signature inspection requires both
transformations in translating a specific mode emission measurement to a
mean emission value. To arrive at suitable relationships between the
various mode emission, parameter settings and mean emission levels
necessitated the partitioning of the engine system into several classif-
ications. For analytical purposes, this partitioning can be expressed
in matrix notation where the columns represent power train types and the
rows, engine subsystems. Depicted below is a symbolic representation of
such a matrix for one emission specie.
~e. =
J
Idle
Ignition
Induction
Air Reactor
all
a21
a31
Power Train
Engine
Modification
a12
a22
a32
The a. ,IS are the response coefficients derived from the orthogonal test
lJ
for each ordered pair of subsystem and power train types. Mathematically,
this can be expressed by the following relationship:
~e. = La.. ~P.
J lJ 1
(4.2)
A similar relationship exists between the several inspection signature
mode emissions and the corresponding engine parameters:
~e = Lb.. ~P.
m lJ 1
(4.3)
55
-------
The partitioning of the engine system i.nto specific subsystems allows for
a more detailed characterization of the contributions made by maladjust-
ment of the various parameters to emission levels. A schematic of the
emissions predictor model is shown in Figure 4-2. With specific policy
criteria as input, the model first ascertains the number of vehicles fail-
ing the given inspection procedure. This is accomplished by projecting
the various pass/fail criteria into the appropriate mode emission or engine
parameter distribution. Except for the air cleaner parameter, all of the
engine parameters evaluated were assumed to be normally distributed. This
assumption leads to a straightforward method for relating pass/fail cri-
teria to the percentage of vehicles whose parameter settings were outside
the criterion range. In the case of the air cleaner parameter, an expon-
ential distribution was found to be a good fit of the experimental test
data. Emission distributions are positively skewed and, therefore, can-
not be described in terms of a normal function. Thus, logarithmic trans-
formations of the emission data were required resulting in a log-normal
distribution characterization.
Using the inspection procedure policy criteria, the model divides
the population into two fleets--an accepted and a rejected subfleet. The
rejected vehicles will undergo maintenance resulting in a reduction or
increase in overall emission levels. The mean emission levels represented
by both subsets are then deteriorated over the subsequent inspection
interval. It should be pointed out that all emission levels and parameter
values utilized in this analysis are in terms of a mean value which is the
composite of accepted and rejected subfleets. This characterization per-
mits a number of simplifying assumptions, in particular in the case of
adjusting emissions levels and parameter values for the effects of main-
tenance treatments. Although the variability of engine parameter adjust-
ments is assumed constant with time, the mean values are adjusted after
each maintenarr.e interval.
Due to lack of sufficient experimental data on both emission and
parameter deterioration rates, the model assumes that all engine para-
meters within a subsystem decay identically and according to the linear
rule:
56
-------
POLICY ACCEPTE D EMISSION
POPULATION
DECISION - -~-," - -- -- -- DETERIORATION
EMISSION LEVEL
. INSPECTION
PROCEDURE
WEIGHTED
. INSPECTION POPULATION
TYPE EMISSION
AVERAGE
. MAINTENANCE
TREATMENT
(J1 REJECTE D REINSP~CI10N
"'-J . PASS/FAIL POPULATION FOR EMISSION MAINTENANCE EMISSION
CRITERIA . AIR SIGNATURE TREATMENT DETERIORATION
. EM APPROACH
Figure 4-2. Emission Predictor Model Flow Diagram
-------
dPi - de./dM
K J
dM - de./dP.
J ,
(4.4)
where:
M = mileage
K = scaling constant proportional to number of parameters
within a subsystem
P. = ith engine parameter mean value change with mileage
1
e. = jth composite emission
J
The subsystem emission deterioration rates in Table 3-7 (de./dM) are used
J
directly in computing the changes in composite emission levels as a
function of time. The response coefficient, dej/dPi' is derived from the
statistically designed experiments (see Volume III, Section 4). After
the deterioration process has taken place, the two subfleets are re-
combined to obtain an updated mean emission level. This recombination
is based on emission levels from both subfleets as well as the inspec-
tion rejection fraction. Equation (4.5) presents the relationship between
these components.
e. = reR' + (l-r) eA'
J J J
(4.5)
where:
ej = jth emission for combined fleet
eRj = jth emission for rejected fleet
eAj = jth emission for accepted fleet
r = rejection fraction
In the idealization of the model, all of the above events occur
instantaneously at the end of an inspection interval. The adjustment
deterioration transient over the next inspection interval for the
accepted and rejected vehicles is taken as the extension of the previous
decay transient. The deterioration profile is shifted by an appropriate
degree to account for differences in absolute emission levels attributable
to maintenance. At the end of an inspection interval, the model combines
the partitioned test population and computes a weighted average mean
emission level. The test population then undergoes another instantaneous
inspection and maintenance process, followed by subsequent inspection
intervals in a manner completely analogous to that described above.
58
-------
To obtain a measure of the effectiveness of the various procedures,
it was necessary to compare their performance wi th some 'Ibenchmark. II For
this study the fleet undergoing enforced maintenance was contrasted with
a fleet subjected to voluntary maintenance reflecting the ARB surveillance
data. A model was developed for predicting the emission levels of this
baseline fleet over the four-year program. The maintenance treatment pro-
gram for the baseline consisted of a yearly minor adjustment and ignition
repair with a major tuneup every third year. To allow for the situation
where the enforced maintenance treatment is less than the baseline, a
Ilconservation of maintenance" principle was applied. The principle states
that the sum of all subsystem maintenance is invariant with time, al-
though that maintenance which occurs as a consequence of inspection is
more effective. The vehicles which pass the inspection procedure are
assumed to get the same voluntary maintenance as the baseline fleet.
All of the above is directed at predicting emission histories which
must then be transformed into a single valued figure of merit. This
implies performing an integration of emissions for both the test and base
populations using the predicted emission histories. For the base popula-
tion, this involves integrating each segment and summing over all segments.
In the case of the test population, however, the emission profiles for
both the accepted and rejected vehicles must be integrated separately and
then weighted to obtain a mean integrated value for that inspection
interval. These weighted average values are then summed over all inspection
intervals and subtracted from the equivalent base population total to
obtain the gross value of reduced emissions effected by the imposed pro-
gram. All of the emission computations performed within the model are
done in terms of composite measurements. However, the emission time
histories and total daily emission levels are converted to grams per mile
and tons per day, respectively. These conversions are done using pro-
cedures set forth in the Federal Register, Reference 9.
The major assumptions and ground rules utilized in formulating the
emissions predictor model are summarized below:
.
All emissions and engine parameter transformations are based
on mean values.
All inspection/maintenance activities are performed instant-
aneously.
.
59
-------
. Engine parameter adjustments and emission levels are
deteriorated at the subsystem level.
. Conservation of subsystem maintenance is imposed.
. Vehicles are maintained to manufacturer's specification
when they fail an inspection.
. A 100% reliability is achieved in detecting engine para-
meter malfunctions.
. Inspection of parameter adjustments at the garage prior
to maintenance is required of all vehicles failing an
emission inspection.
. The program is evaluated over a four-year period utiliz-
ing 1966-69 control technology.
. Subsystem parameter adjustments and emission levels decay
linearly with time.
. Average driving speed to and from inspection stations is
22 mph.
. The transformation between mean emissions and log-normal
mean emissions is linear.
4.2 COST ESTIMATOR MODEL
The economic elements of the system are highly sensitive to both the
inspection frequency and specific engine parameter and/or emission mode
pass/fail criteria. In addition, economic considerations are closely
related to the number and type of inspection and maintenance facilities
available in the demographic area being simulated. In general, the
economics of an imposed inspection and maintenance program involve both
direct and indirect resource costs. The former are the common costs
related to the real exchanges of funds for material, goods and services,
while the latter include resource expenditures that are not necessarily
accompanied by real flow of funds.
The costs associated with capital investment requirements are a
function of the basic inspection/maintenance strategies. For the case
of a franchised-garage system, no direct capital investment will be
accounted for in the program. This assumption presupposes that fran-
chised licenses will be awarded only on condition of a satisfactorily
equipped garage. The state inspection/franchised-garage maintenance
strategy does require a capital investment for the inspection facilities.
60
-------
These costs will depend on the size of the car population, the length of
the inspection interval, and specific instrumentation and equipment re-
quirements. In the model, it is assumed that all investments come on-
stream at time zero and that no existing state facilities are employed.
Separate investment calculations are performed for the building, land and
equipment requirements for each state-lane station. The contribution of
these investment costs to annual costs is calculated using the concept of
a sinking fund. This can be interpreted either as financing through
bonds which are payable in full upon maturity or through internal funds
which must be replaced in full at the end of a fixed period.
For each inspection interval, direct operating costs must be added
to the above indirect operating costs to obtain the total system operat-
ing costs. Under the direct category are separate calculations for wage
costs, administrative costs and miscellaneous operating costs for both
inspection and maintenance. In addition, a charge for parts is incurred
under the maintenance activity.
The system costs so far discussed are explicit costs which involve
real exchanges of funds for material goods and services. From a social
standpoint, however, the costs that result from such an inspection and
maintenance program are better measured by the appropriate costs of all
resources employed in the program; i.e., the value of all resources if
employed in their best alternative use. If it is assumed that the data
employed in the above calculations are rough measures of the marginal
values of those resources employed in their best alternative uses, it is
necessary to add an estimate of the implicit cost represented by driver
inconvenience to the computed explicit costs to arrive at a better approxi-
mation to the system social cost. The total time spent by users in inspec-
tion and maintenance is considerable, and it is reasonable to expect that,
without an imposed program, this resource would be employed in other pur-
suits from which the individual would derive a benefit. The model, there-
fore, computes a monetary estimate of these lost benefits which compose
the value of private time expended on the program.
61
-------
The model also includes procedures to account for time effects on
system costs. It is desired to base the cost calculations on inflated
rather than constant dollar costs, all costs can be inflated at a con-
stant annual rate. Because of available mechanisms for employing funds
productively and, thus, earning an interest or profit, it is not realis-
tic to weight a dollar cost incurred today on an equal basis with a
similar cost incurred at some future date. Future costs must be dis-
counted by a rate related to the productivity of capital. All costs are
thus discounted at an assumed constant annual rate.
In summary, once system investment costs have been computed, the
model evaluates the inspection interval, indirect and direct operating
costs, and the private user-costs for both inspection and maintenance.
These costs are summed for a single interval, discounted by factors
appropriate to the end of that interval and then summed over all intervals
to provide a measure of total program cost. Figure 4-3 depicts the re-.
lationship between the various cost components comprising the system. The
flow diagram was constructed for the case of a state-lane inspection/
franchised-garage maintenance strategy. For a franchised-garage system,
no capital costs would appear.
4.3 OPERATIONS MODEL
The operations model provides the linkage between the emission pre-
dictor model and the cost estimator model. It calculates the size,
location and number of inspection and maintenance stations required to
implement the program optimally. In making these determinations, the
model trades off user-inconvenience costs with the capital costs for
facilities. It does this by computing the driving time to and from the
station and waiting time at the station. Waiting times at inspection
station locations are derived from a queuing model. This model assumes
a Poisson distribution of vehicle arrival times at the inspection station.
Adding the traveling time and waiting time to the actual vehicle inspection
time yields the total inconvenience time. For the case of the emission
signature inspection, an added travel time for some vehicles is incurred
for reinspection. This additional time is charged against the program if
the failed vehicle passes its second inspection. No user time charges are
62
-------
CTI
W
INSPECTION
L:
. ANNUAL COSTS
. COST PER
INSPECTION
. COST
BURDEN
. TOTAL
PROGRAM
COST
TOTAL
OP. COSTS
FOR
EACH
YEAR
I
I
I
I
,..------, I
INCONVENIENCEl- ---- -_--J
L_____.J
MAINTENANCE
CAP: CAPITAL
OP: OPERATING
Figure 4-3. Cost Estimator Model Flow Diagram
-------
allocated for the corresponding maintenance activities since these would
probably have to be performed even in the absence of an enforced program.
The conversion of these times to a direct dollar amount is done using
a social cost of $l/hr. Thus, a direct comparison can be made between
these costs and the designed facilities costs. The more stations deployed
in the system, the lower the social cost to the public. However, capital
investment costs required to set up the system may more than offset the
savings achieved by the shorter driving times.
Another operational aspect analyzed by this model is the tradeoff
between a labor-intensive versus an equipment-intensive system. Here,
the implications of a fully automated monitoring and recording system
are traded against the inherent advantages of a manual, low capital equip-
ment system. Generally speaking, the equipment costs associated with the
measuring and recording equipment are small in comparison with the overall
operational costs and, thus, the issue revolves around the reliability of
each system.
4.4 CRITERIA FOR POLICY EVALUATION
Up to this point, we have described the manner in which the simula-
tion model computes measures of emission levels and relevant costs for a
given policy statement regarding vehicle inspection and maintenance. We
need now to identify an acceptable figure of merit which embodies within
a single value a measure of the goals of the program. Comparison of dif-
ferent values of the figure of merit permits rapid assessment of relative
economic-effectiveness of various policies. Ideally, then, we would select
that policy set (i.e., inspection interval, emission level, pass/fail
criteria, etc.) with the "best" figure of merit. Here, the term "best"
refers to the lowest value of the figure of merit.
Unfortunately, no one figure of merit appeared to embody all of the
desired characteristics. The simulation model instead has the capability
of examining several different figures of merit, thus permitting a deter-
mination of the sensitivity of the optimal decision to the chosen objective
function. The most relevant figure of merit and the one utilized through-
out this study can be expressed by:
64
-------
F. f M .t - Program Cost
19ure 0 erl - EW. ~E .
1 Cl
where:
Wi = weighting function for each emission specie
~E . = emission difference between baseline and test program
Cl
This relationship provides a basis for comparing base and test population
emission levels with the cost of the test program. The figure of merit
units are in discounted dollars per weighted tons of emission reductions.
Thus, program effectiveness can be read in terms of so many dollars to
achieve a reduction of one composite ton.
As can be seen, the weighting function establishes the degree to
which emission specie reductions impacts the program design. Having
fixed the weighting of emission reductions the model can determine the
optimal pass/fail criteria and system design for the several proposed
inspection/maintenance strategies. In actuality, the weighting of emis-
sion reductions must reflect the air pollution problem of the various
urban centers. Since some regions are more concerned with high ambient
CO levels than with HC levels, they would weight CO reductions higher than
those of HC or NO. As discussed in Section 2, the Los Angeles Basin was
selected for testing the feasibility of a program of vehicle inspection/
maintenance. For this case, the weighting function given in Table 4-1 was
utilized in the overall economic~effectiveness evaluation.
Table 4-1.
Regional Weighting Function
Emission Weight
HC 60%
CO 10%
NO 30%
This weighting function, developed from EPA data (Reference 10) places
equal weight on both HC and CO in terms of actual tons reduced. That is,
since CO emission levels in terms of tons/day are six times those of HC,
the net contribution of both after weighting is about the same.
65
-------
The mechanism for actually determining the most attractive policy
set from among the several proposed inspection/maintenance approaches
should include not only the figure of merit, but also the inspection costs
per car and attendant emission reductions. These latter two parameters
are important in that they relate to the practical aspects of emission con-
trol. For example, it could be assumed that a program producing emission
reductions less than 15%, no matter how economically attractive, would
not be very effective as a control scheme. The guidelines selected for
this study are in the form of the following cost constraints and emission
reductions goals:
. Average cost of six dollars per car was the
. A program which provides emission reductions
of less than 15% was unacceptable.
maximum allowable
(HC and/or CO)
These guidelines when used in
the criteria for analytically
and system design.
conjunction with the figure of merit provide
identifying the "best" inspection procedure
Each combination of inspection procedure and system design can be
looked upon as a strategy. By examining the various strategies with the
economic-effectiveness model, an ordinal ranking of these strategies
based on their figures of merit can be developed. It becomes fairly
straightforward to then identify the optimal strategy by merely selecting
the one which ranks first and conforms to the cost and actual emission
reductions guidelines described above.
66
-------
5.
MODEL SIMULATION RESULTS
The preceding sections described the system framework, assumptions,
and input data required to evaluate the economic-effectiveness of an
inspection/maintenance approach to vehicle emission control. Two basic
questions to be resolved are: 1) the viability of the approach, and 2)
the most attractive procedures in terms of a stated criteria.
The answer to the first question lies with the establishment of
specified emission reductions goals and associated economic constraints.
For this study emission reductions of approximately 15% for a specified
emission specie(s) were required of each maintenance treatment in order
to consider the procedure set viable. For example, the idle adjustment
program is designed to primarily reduce CO emissions. Therefore, any
idle maintenance procedure should provide at least a 15% reduction in CO
emissions. Furthermore, any selected procedure should not increase the
hydrocarbon concentrations above the stated baseline. As for the cost
implications, it was held that for a program to be feasible the average
cost should not exceed $6/car/inspection interval. These two criteria
(i.e., 15% reduction and $6 maximum cost) when used in conjunction with
the figure of merit formed the criteria for selecting the most cost-
effective system.
Two basic inspection/maintenance approaches were examined within this
context.
Engine Parameter Inspection: This approach provides for the direct
inspection of the various parameter adjustments specified in the procedure.
If the inspection procedure reveals parameter settings outside a pre-
determined range they are reset through appropriate maintenance to manu-
facturer's specification.
Emission Signature Inspection: Under this alternative, measurements
are made of the vehicle's exhaust emissions using several engine loadings
(i.e., driving modes). If the resultant mode emissions are higher than
some preset criterion, the vehicle will be rejected and will undergo direct
diagnosis at the repair faci1ity and, if required, corrective maintenance.
67
-------
Several mode emission signatures have been identified that allow detection
of component malfunctions at the subsystem level.
Embodied in each of these two approaches are a number of substrategies.
Simulations were performed to evaluate the degree to which the figure of
merit, and, therefore, the policy decision is sensitive to changes in the
values of the design variables. For the case of multidecision inspection
procedure (i.~., several parameter settings or mode emissions) a matrix of
candidate values was evaluated by holding all but one variable constant.
The analysis was conducted in two phases. The objective of the first
phase was to evaluate the various alternatives within each of the major
strategies. Here, tradeoffs between maintaining engine subsystems (i.e.,
idle, ignition and induction) and combinations thereof were performed.
In order to obtain a consistent base for comparing the effectiveness of
the subsystem strategies, an optimization of procedures was undertaken.
In the case of the engine parameter inspection procedure, this involved
determining optimal pass/fail values for timing, rpm, idle CO and air
cleaner blockage. For the emission inspection procedure, emission signa-
tures using both idle and loaded CO and He modes were evaluated. Based on
a consistent evaluation of this information for both basic strategies, an
assessment of the general feasibility of a vehicle inspection/maintenance
program was made.
Presented in Table 5-1 are representative summary results which fit
our performance criteria for the several strategies examined with the
model. Depicted for each engine parameter inspection and emission inspec-
tion procedure are the figure of merit, cost per inspection/maintenance
per car, and expected emission reductions. Each strategy presented has
been optimized with respect to the figure of merit by varying the pass/fail
criteria for the engine parameters or mode emissions and the interval of
inspection. Inspection of Table 5-1 reveals that the most cost-effective
procedure for reducing CO is an engine parameter inspection followed, if
necessary, by an idle maintenance treatment. Within this option two
inspection alternatives are evaluated, a state-lane inspection and a
franchised garage inspection. Applying our emission reductions goals
criteria it would appear that either procedure will be satisfactory.
68
-------
Table 5-1.
Summary of Most Cost Effective Inspec~ion/Maintenance Procedures
4-Year Average Emission Reductions
0'\
\.D
Cos t P e r E mi s s i 0 n Re du ct i OIL -
Fi 9 u re 0 f Me ri t Veh i c 1 e H C CO N 0
I ns pe cti on / Ma i n te n a n ce P ro ce d u re $/ To n $ % I
E n gi ne P a rame te r Di a gn os t i cs
A . S h 0 rt I n s pe cti 0 n ( S ta te L an e )
. I d 1 e ( I CO an d rp m) I:::::::::::::::::::::::::::::::::et:~::::::::::::::::::::::::::::::::::::::::::::::':::: ::::::::::::::::::::,,::::::::::"~:::~::I,:::::::::::':::::':,:::,:,::,: ::','::':"::'::::::':::':::::::,::::::!:::~::::::::!::::j!!!j!j!~!!:::::::::::::::::::::::::J:t:::::::::::::::
I
B . L 0 ng I ns pe ct i on ( F ra n ch i s e d )
. I dl e ( I CO , rpm t i mi n g ) 366 2 . 50 3 1 3 - 3 I
,
. I d 1 e + I gn i ti 0 n ::::::::::::::::::::::::I:::::~~B::::::::::::::::::::::::::::::::::::::::I::::::: ::::::::::::::::::::::::::::::::~:~:H9:i:::'::::::::,::::::i:i::::::: ,::::::::::i:::::::i,:,i,i::::::::::::~:i:::i::::::::::j:j:::,:f::::::i:::::::::i::::::::j:j:::I:j:j:j::::::::::~
. I dl e + I 9 n i t i 0 n + I n d u cti on 5 3 7 1 2 50 2 2 33 - 5
E mi s s i 0 n S i gn a tu re An a 1 y s i s
A . I d 1 e Mo de ( S ta te L a n e )
. I d 1 e ( CO emi s s i on ) 42 2 2 . 50 2 1 2 - 4
. I d 1 e ( CO a n d H C emi s s 1 0 n ) 4 5 8 2 . 50 3 1 1 - 4
B . Lo a de d Mo de ( S ta te L ane )
..............."........".".".,......_----,-'.,.. """""""""""'''''''''''..,.'''" ",'."[[[
.........,-,..,-,.,..,-,......................,.. ::::::::::::::::,:::::::::::::':~:;:I:::::::::::::::::Ii::::::::::::: [[[-
"""""""""""""""3.5~f""""""""""',',',",',',',', ',',',',' """"""""""""lr"""""""''1~'"''''''''"""""""'81("""""'"
I d 1 + I i t i ................. "'"'''''.''''''''''''''' . . . - . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
e gn on ..............,.. . .......................... . . . . . . . . . . . . . . . . . ... . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . ..
. ......,-,,-,...,- ..,....................... ......."""""'''' ............ ............... .........
,~ftjj??~\\\::i:~::::~}\}\}}}rrrr}f~~ ~~~~~~~~~~;~~~~~~~~~~~~~~~~~~~~~~~~~~~~::~~:::~~~~~~~~~~~~~~~~~~~~~~~::~::;::~~~~~~~~~~~~~~~~~~~~~~~~~~::::::::::~t~~t~~~~~
,',',',', """"""""""''4ff'"'''''''''''''''''' ,',',',',',',',',',', "',', """""""""""""""p""'O()X""""""'"',',',' ,:::,:::.:,:::::::,:::::::::j:::::i:~i:::::::::::::,:j:::t~::j:::::j::::::::::::::::;:I::::j:j,j:::::::
I d 1 I i ti + I d ct i ................... """'''''''''''''''''''' .................... . ................
+ gn 0 n n u on ................. .. """"""""""""'" .................... ...............
. e ................. ... .......................... ................ ... . ................
-------
Actually, if more emphasis were placed on reducing CO levels, i.e.,
through a reweighting of the payoff function, the state-lane procedure
would emerge as the most cost effective. Examining the more elaborate
engine parameter inspection/maintenance procedures (i.e., idle + ignition
and idle + ignition + induction) indicates that substantial reductions in
emissions can be achieved. However, these reductions are more than off-
set by increased program costs, primarily attributable to repair parts
and increased inspection/maintenance labor times. The cost of the latter
procedure, actually exceeds the upper economic constraint established for
the study. NO emissions increase because of the higher combustion temper-
atures associated with leaner carburetion. This increase is partially
offset by retarding the basic timing.
In order to determine the long-term effects, it is important to
examine the emission time histories for the various procedures. Shown
in Panels A, Band e of Figure 5-1 are He, CO and NO emission profiles,
respectively, for the idle engine parameter inspection/maintenance pro-
cedure. For this analysis, the identification of acceptable procedures
was limited to those that substantially reduced He and CO emission levels.
The main conclusion to be drawn from Figure 5-1 is that an idle adjustment
program has its largest effect on CO levels. While an average reduction
figure of some 13-15% was obtained from this procedure, it is obvious that
a much larger CO reduction, i.e., 20-25% can be achieved once an equil-
ibrium point has been reached. The equilibrium point is defined as that
emission level where the fraction of vehicles being rejected is a constant
and the emission level has stabilized at a new level. For CO an equil-
ibrium point at year four has not been reached and, therefore, the anti-
cipated emission reductions for year five should be greater. The small
differences in baseline and test fleet He emission levels can be attri-
buted to the slight effect of the idle parameter adjustments (e.g., idle
CO, rpm, timing) on composite He. A.small effect also is shown in in-
creased NO emission levels. For all the alternatives examined, the NO
emission level did not increase beyond 7% and usually remained invariant
around 4%. The NO emission profile is most affected by a timing adjustment.
70
-------
5
~
"'-.
0 4
l
V)
...J
UJ
> 3
UJ
...J
Z
0 2
v)0
V)
~
UJ
U
J:
0
~ 40
"'-.
o
l
V) 30
...J
UJ
>
UJ
...J 20
Z
o
~ 10
~
UJ
o
u
~ 8
"'-.
o
l
V) 6
...J
UJ
>
UJ
...J 4
Z
0
V) 2
V)
~
UJ
o 0
Z 0
Figure 5-1.
IDLE ADJUSTMENT PROGRAM
PANEL A
HC
-- ---
~
BASE LINE FLEET
- -- COMPOSITE TEST FLEET
PANEL B
o
CO
~ ----- ---
",. ----
BASE LINE FLEET
--- COMPOSITE TEST FLEET
PANEL C
NO
- ---- -----
-----
-
BASE LINE FLEET
--- COMPOSITE TEST FLEET
2 3
TIME"" YEARS
4
5
Emission Time Histories - Engine Parameter Inspection
71
-------
The model also predicted a similar set of emission time histories
for the emission signature inspection procedure. Presented in Figure 5-2
are representative emission histories for maintenance treatments involving
either an idle + ignition maintenance or an idle + ignition + induction
tuneup. As could be expected, these inspection/maintenance procedures
are more effective in reducing emission levels than for the case of a
simple idle adjustment program. As exhibited in Table 5-1, these two
procedures are the most cost-effective of any of the major maintenance
procedures examined. Their biggest contribution over the idle mainten-
ance strategy is in reducing hydrocarbon emission levels, the primary
reason being the replacement of ignition subsystem components. Although
the emission reductions experienced with an emission inspection procedure
are not as great as with the engine parameter inspection, the overall
cost-effectiveness is appreciably higher.
The third strategic variable analyzed by the model is the frequency
of inspection. Here the interest is in examining the impact of varying
inspection frequencies on program effectiveness, i.e., the tradeoff of
benefits obtained from shorter inspection intervals with the added cost.
A typical tradeoff of inspection frequency for one policy alternative is
presented in Figure 5-3. As can be seen, the lowest figure of merit
value occurs for inspection intervals in the range of 12-14 months. For
administration reasons a 12-month frequency of inspection should prove
most cost-effective. The dotted segment of the curve describes the situa-
tion where the emission reductions approach zero and the figure of merit
infinity.
This discussion has centered on some of the summary level results
obtained from the study. A more detailed discussion of the two basic
inspection policies follows.
72
-------
~ 40
o
~
::J 30
L.U
>
, L.U
-J 20
Z
o
V1
V1 10
:E
L.U
o
U
:E 8
6
I
V1 6
-J
L.U
>
L.U
-J 4
Z
0
l/) 2
l/)
:E
L.U
0
Z
5
~
"-..
0 4
I
l/)
-J
W
> 3
w
-J
Z
0 2
l/)
l/)
:E
w
U
I
0
IDLE + IGNITION AND IDLE + IGNITION + INDUCTION
ANNUAL ADJUSTMENT PROGRAMS
PANEL A HC
~---------- -------- --------
-
-
I
BASE LINE FLEET
-
------ IDLE PLUS IGNITION
---IDLE PLUS IGNITION PLUS INDUCTION
PANEL B
CO
-----
---
o
PANEL C NO
....::..- - -- - - - -------
-------
o
o
2 3
TIME, YEARS
4
5
Figure 5-2.
Emission Time Histories - Emission Inspection Procedure
73
-------
600
500
Z 400
o
~
"--
--
l 300
'-J U.
-t:::> U.
0
>- 200
~
Cl..
100
'- '
~ ~
,
'" t I
" I I
i
.' ,
1"-0... ' ~.
-----'--
'" ~ /,
,
ENGINE PARAMETER INSPECTION,
IDLE ADJUSTMENT MAINTENANCE
2
4
6
8 10 12 14 16 18
INSPECTION FREQUENCY, MONTHS
20
22
24
o
o
Figure 5-3.
Impact of Inspection Period on Program Effectiveness
-------
5. 1
ENGINE PARAMETER INSPECTION/MAINTENANCE PROCEDURES
The engine parameter inspection procedure offers the advantage of
minimizing the number of diagnostic errors (i.e., errors of omission and
commission) during the inspection activity. This characteristic makes
this approach extremely attractive in terms of effectiveness and reli-
ability. Shown in Figure 5-4 are resulting figures of merit obtained in
simulating various pass/fail criteria and combinations of engine idle
parameters. Panels A and B depict figure of merit contours for various
idle CO, timing and rpm pass/fail cutpoints. Two rpm cutpoints, -75 rpm
and -50 rpm, were sufficient to detect substantial changes in the shape
of these cost-effectiveness contours. The more compact the contours are,
the more optimal the system has become. The most cost-effective contour
plotted (i.e., $350-400/ton) suggests that the optimal pass/fail criteria
for the idle subsystem are idle CO = 3.7%, timing = 7 degrees, rpm = 75.
Changes in idle CO are indicated to have the largest overall effect on
program cost-effectiveness even though this emission species has the
smallest weighting factor.
Panel C shows the simulation results for the air cleaner component
of the induction subsystem. The abscissa value represents the degree to
which the air cleaner passages were blocked. An optimal pass/fail value
of roughly 1050 of blockage was determined from the analysis.
These four engine parameters were the only ones optimized with
respect to a range of cutpoint values. The other three engine parameters--
ignition, air pump and PCV--were maintained on the basis that they failed
a single cutpoint as discussed in Section 3. For the PCV system, all
vehicles (approximately 12%) which had positive crankcase pressures were
assumed to be sufficiently plugged to require repair. Any vehicle mis-
firing in the range of engine load of the seven-mode cycle were also
failed. These assumptions are consistent with the data acquired. How-
ever, for the case of the malfunctions involving air-fuel flows such as
PCV and air reactor system restrictions, a parameter optimization would
be possible if vehicle population distributions of flow rate through
these devices were available. Since this was not the case, it was assumed
that only those devices which were completely failed were maintained.
75
-------
5.0
4.0
3.0
-;R.
o
o
u
W
...J
o
5.0
4.0
3.0
700
z
o
I-
"-
""
!::=
c><:
:g 600
u...
o
w
c><:
::>
o
u...
500
o
Figure 5-4.
I
FIGURE OF MERIT
S/TON
350 - 400
----- 400 - 450
--- 450 - 500
PANEL A - IDLE ADJUSTMENTS.6 RPM =-75
---....
,-
,
I
I
I
\
,
\
,
o
2
4
6
8 10 12 14
/:. TIMING - DEG
PANEL B -IDLE ADJUSTMENT.6RPM =-50
20
40
60
80 100 120 140
BLOCKAGE - DEG
PANEL C - AIR CLEANER
16
18
22
20
160
180
220
200
Engine Parameter Subsystem Optimization
76
-------
Table 5-2 presents an optimal set Qf engine parameter settings for the
several subsystem combinations. These criteria produce a series of emis-
sion time history plots similar to those presented in Figure 5-1. Where-
as, Figure 5-1 related emission histories for an idle inspection/mainten-
ance program, Figure 5-5 relates emission histories for an idle + ignition
and idle + ignition + induction tuneup. As anticipated, these more
complete maintenance programs reduce emission levels substantially below
those produced by an idle adjustment program. This is especially true in
the case of hydrocarbon emissions. In all cases these programs tend to
improve with time. This condition is attributed to the fact that the
vehicle population has not reached a state of equilibrium.
The simulation model examined the operational characteristics of the
various inspection/maintenance alternatives over a four-year time horizon.
Selection of a four-year period was based on the desire to minimize the
impact of the following factors on determining the most cost-effective
program design.
.
Uncertainty in predicting the implications of proposed 1975
Federal standards.
Uncertainty in estimating the effects on emissions of major
engine repair beyond four years.
Lack of experimental emissions deterioration data beyond
50,000 miles (i.e., approximately four years).
.
.
As stated earlier, the vehicle population fleet was divided into two
emission control c1asses--air injection reactor (AIR) and engine modifica-
tion (EM). Figures 5-6a, band c presents the emissions profiles for these
power trains for an idle adjustment program. Hydrocarbon emissions show
that the AIR emission controlled vehicles are slightly more sensitive and
responsive to an idle adjustment program. The higher sensitivity of AIR
cars to the various maintenance programs can be traced to both the larger
emission decrements achieved as well as to their lower rates of emission
decay. Data utilized for this study showed that AIR cars as a general
class have higher emission levels initially than do engine modification
vehicles. Since most AIR cars are of an older vintage, (i.e., 1966-67)
their emission levels would tend to be higher due to the additional mile-
age accrued. The main reason that the weighted average emission curve
77
-------
Table 5-2.
Optimum Inspection Strategies 4-Year Average Emission
Reductions
Weight Factors: HC = .6, CO = .1, NO = .3
ENGINE SUBSYSTEM
OPTIMUM EMISSION INSPECTION
MODE EMISSION CUT POINTS
. IDLE, FUEL TO AIR ADJUSTMENT ONLY
ICO = 4.0%
. IDLE, IDLE PARAMETERS
ICO = 4.0%
IHC = 300 PPM
. IDLE + IGNITION
OPTIMUM ENGINE PARAMETER
ICO = 2.5%
IHC = 300 PPM
LHC = 400 PPM
ICO = 2.5%
IHC = 300 PPM
LHC = 300 PPM
LCO = 1.0 %
INSPECTION
. IDLE + IGNITION + INDUCTION
ENGINE PARAMETER
. AIR CLEANER
PARAMETER CUT POINT
105 DEG. BLOCKAGE
. PCV
. AIR PUMP
PLUGGED
FAILED
. MISFIRE (IGNITION)
. TIMING
2.5%
. RPM
7 !::, DEG.
-75 !::, RPM
. IDLE CO
4%
78
-------
ENGINE PARAMETER DIAGNOSTIC
EXTENSIVE MAINTENANCE
5 PANEL A
~ HC
"
l') 4
l
V1 ---
-I 0--.-- --- --
u.J 3 '-'-. --
> .--.
u.J
-I
Z
0 2 BASE LINE FLEET
V1
V1
~ ----IDLE + IGNITION
u.J
u -'-IDLE + IGNITION + INDUCTION
:c
0
~ 40
"
l')
l
j 30
u.J
>
u.J
-I 20
PANEL B
CO
---
NEW CAR
EMISSION.-
LEVEL
'--.
'--
0---._._.
Z BASE LINE FLEET
o
V1 ----IDLE + IGNITION
V1 10
~
u.J
o
u
-'-IDLE + IGNITION + INDUCTION
o
~ 8
"
l')
l
V1 6
-I
u.J
>
u.J
-I 4
Z
0
V1 2
V1
~
u.J
0
Z
PANEL C
NO
- -- -=-===-- -
~:::--
BASE LINE FLEET
----IDLE + IGNITION
-.-IDLE + IGNITiON + INDUiTION
00
2 3
TIME"'" YEARS
4
5
Fi gure 5-5.
Emission Time History for Several Parameter
Inspection Procedures
79
-------
c.n
W
...J
~
"""
c.n
~
6£
(9
~
c.n
...J
W
>
OJ UJ
a ...J
Z
o
c.n
c.n
~
UJ
PAN E L A
ENGINE PARAMETER DIAGNOSTICS
IDLE MAINTENANCE PROGRAM
5
HC EMISSIONS
4
--
--
.....""'" . -. --- .
......... ....................... .......................
3
2
......... ....
----
---
1 0
1
2
TIME - YEARS
--
--
-
_0
. -----. ........ .
BASE LINE
AIR
CAP
AVERAGE
3
4
Figure 5-6a. Power Train Emission Histories - Panel A
-------
PAN E L B
ENGINE PARAMETER DIAGNOSTICS
IDLE MAINTENANCE PROGRAM
50
CO EMISSIONS
40
UJ
....J
~
"'"
V) -....
~ 30
<{
0:::
0
l
V)
....J
UJ
co >
--' UJ 20
....J
Z
0
V)
V)
~
UJ
10
--
BASE LINE
----- AIR
--- EM
---- AVERAGE
G
o
1
2
TIME - YEARS
3
4
Fi gure 5-6b.
Power Train Emission Histories - Panel B
-------
V1
W
.....J
~ 6
"---
V1
~
<{
0<::
0
~
V1 4
OJ .....J
W
N >
W
.....J
Z
0
V1 2
V1
~
W
PANEL C
ENGINE PARAMETER DIAGNOSTICS
IDLE MAINTENANCE PROGRAM
8
NO EMISSIONS
----.
--.--
.-. r-.-.-.-.-
----- --- .-.-.-.-
--
-- 1-00-._------
--- -- ---
.. ........... ... ............ ........................
....... ....... ....... ......... .. .........
BASE LINE
............ AIR
-.- CAP
---- AVERAGE
o
o
1
2
TIME - YEARS
3
4
Figure 5-6c. Power Train Emission Histories - Panel C
-------
does not more strongly reflect this fact is that AIR vehicles comprise
only 30% of the Los Angeles population. CO emissions are nearly equally
responsive to idle adjustment for both power train types, Panel B. NO
emissions show (Panel C) a trend of a gradual reduction over time. This
is largely attributable to richer carburetion (i.e., increasing CO emis-
sions level) with time when enforced carburetor maintenance is not
imposed. AIR equipped vehicles are less sensitive to F/A ratio since CO
emissions are oxidized to C02 in the air reactor, hence, AIR vehicles
are operated at richer carburetion with lower NO emissions.
A summary level matrix of engine parameter diagnostic procedures is
presented in Table 5-3. Examination of this table reveals that the short
inspection-idle maintenance procedure is most cost-effective. However,
the resultant HC emission reductions are not appreciable, ranging from 0
to 3%. The basic difference between these two candidates is that the
first one employs a state-lane inspection system whereas in the second
both the inspection and maintenance activities are conducted in a fran-
chised garage. In terms of the actual inspection process, only idle CO
and rpm are examined in the state-lane system. This situation accounts
for the fact that changes in both hydrocarbons and oxides of nitrogen are
negligible, these species being strongly affected by timing. The average
cost-per-vehicle of the two alternatives falls between $1-3 and is not in
conflict with the $3/car reported for the GM adjustment program and the
$6/car reported for the present New Jersey inspection program. The other
alternatives listed, although producing larger emission reductions either
have significantly larger figures of merit or fail to meet the stated per.
formance criteria.
Although the most complex inspection/maintenance procedures are
highly effective in reducing emissions, costs of both inspection and
maintenance escalate rapidly. For example, every vehicle must undergo
an ignition system diagnosis with an electronic analyzer even though
only 3-5% of the vehicles will be found to be misfiring. This increment-
ally increases the cost per vehicle by $2.50 without, as yet, effecting
an emission reduction by subsequent repair. A similar problem is found
with the induction system where costly inspections are required to find
83
-------
Table 5-3.
Optimal Engine Parameter Subsystem Inspection Strategies
4-Year Average Emission Reductions
--_..
-. _._--~--.~.- -------. ---
WEIGHT FACTORS: HC = .6, CO = .1, NO = .3
SUBSYSTEM INSPECTED
FI GURE OF MERIT
$/TON
COST PER INSPECTION
$
% EMISSION REDUCTION
HC CO NO
. IDLE PARAMETERS
. STATELANE
:.::::~:~::?i:.:.::i::.....::i::::::.:::::.::.i:i[[[i:i:::.:::.:::::::1,::8::~.::::::...::::.::.::'[[[ti~:::.:.:.::..:::...::::::::::.::::.:1.:~:::::.::.:::.::.:"::.:t:7..:::
366 2. 50 3 13 -3
:i:ii!ill?i:::ii:i:i::ii:i::::::::::::::::::i:::::::::::::::::::::::.:::.:::::::::::::i:::::::::::::::::::::::::.:::i::::::I:~::9g:::::::::.::::::::::::::::::i.:::::::I:::i:::::iii::.i:i.:.:::::::::::::~:~:i:::ii::i:::i:i:::::i[[[i~::::::.
:.::1!~!~::i1:1....:::i.i:::!!i:.:i..::.:.:::i[[[1.ii:ii::::.::::::::::::~::i:9g::::::.i:::::::::::.:::.:::.:::::::i:::.:::.:.:.:::.i:i::::.::::::.:.:.:::::.::~:i:i:i::.:::::::::::::::.:.::::::::J::ij::i.::::i:::i::::::::.Sg:i:t
-------
repairable malfunctions which affect air-to-fuel ratio. In addition, the
costs of repair in terms of parts and direct labor charges are substantially
higher than for the simple idle adjustments.
The result is significantly highet' costs to effect emission reductions,
even though maintenance is quite effective on those vehicles with diagnosed
fail ures. Thi s fact suggests that if the inspection costs common to all
vehicles can be substantially reduced, a more cost-effective procedure
will evolve. As we will show in the following section, an emission
signature inspection procedure will partially satisfy this requirement.
5.2 EMISSION SIGNATURE INSPECTION/MAINTENANCE PROCEDURES
Conceptually, a mode emission inspection offers several distinct
advantages over an engine parameter inspection. The advantages lie with
lower cost and relative ease of implementing the inspection procedure.
The basic disadvantage of this strategy is the recurring problem of vehicle
inspection errors. These inspection errors are the direct result of the
inability to ascertain precisely which vehicles have malfunctions using
an emissions measurement. This imprecision is caused by the confounding
effects of engine malfunctions on mode emissions. In some instances a
mode emission inspection will allow vehicles to pass even though they have
engine malfunctions. This type of error is known as an omission error.
Conversely, the testing procedures may fail a vehicle even though it does
not have a malfunction which can be repaired with adequate cost-effective-
ness. This type is called commission error.
The economic-effectiveness model was used to examine a wide range of
emission inspection alternatives. Shown in Figure 5-7 are a series of
simulations for various emission inspection procedures and their associated
cutpoint criteria. An inspection/maintenance program using idle CO as an
inspection criteria is shown in Panel A. The coupled idle maintenance
treatment consists of adjusting the rpm, timing and idle CO. An optimal
value of around 4% gives the lowest figure of merit. Panel B shows the
sensitivity of program effectiveness to variations in hydrocarbon cutpoint.
In this idle emission screening procedure both CO and HC were used as mal-
function indicators. An idle hydrocarbon cutpoint of 400 ppm in conjunction
85
-------
Z 600
o
I-
""
~
~ 400
u..
u..
o
>- 200
~
a...
~ 600
I-
""
~
~ 400
u..
u..
o
~ 200
a...
6 600
I-
""
~
~ 400
u..
u..
o
~ 200
a...
PAN E L A
SUBSYSTEM SENSITIVITY TO IDLE CO CUTPOINT
800
IDLE ADJUSTMENT (I CO)
o
o
4
5
PANEL B
3
IDLE CO ~%
SUBSYSTEM SENSITIVITY TO IDLE HC CUTPOINT
2
800
IDLE ADJUSTMENT SUBSYSTEM
ICO CUT POINT = 4% '"
~ ~
TIMING
RPM
o
o
300
IDLE HC ~ PPM
PANEL C SUBSYSTEM SENSITIVITY TO LOADED HC CUT POINT
400
500
100
200
800
I I
MISFIRE - ELECTRICAL SUBSYSTEM
" ""'" ~
~ ~ ..-.
o
o
300
LOADED HC ~ PPM
400
500
100
200
Figure 5-7.
Sensitivity of the Figure of Merit to Mode Emission
Inspection Criteria
86
-------
with an idle CO cutpoint of 4% yields the optimal figure of merit. Panel
C depicts the results for an idle plus ignition inspection/maintenance
policy. Here, HC emissions measured with the engine loaded in accelera-
tion is used to diagnose the state of the vehicle's ignition system.
This loaded mode which was used in screening both ignition and carburetor
subsystems malfunctions provided the least confounding of any of the modes
analyzed. Again a value of 400 ppm appears to be optimal. Summarized in
Table 5-2 are actual optimal cutpoint criteria obtained from eva1uating
the alternatives within the mode emission inspection approach. These
combinations of cutpoints provide, in general, the best figure of merit.
However, we will see that in some cases these procedures are not compatible
with system performance requirements.
The effectiveness of the various inspection/maintenance alternatives
are indicated by the emission time histories presented in Panels A, Band
C of Figure 5-8. Shown for each specie are emission histories for an idle
emission inspection and adjustment program. The screening procedure used
in this case was a combination of idle CO and idle HC emissions. This
combination of emission inspection modes provided a less effective proce-
dure than when idle CO was used alone. Idle HC was added to detect timing
and rpm malfunctions. As with the engine idle parameter inspection
approach, this program has its largest effect on carbon monoxide reduction.
The general trends of these plots are much the same as those exhibited in
Figure 5-1. The lower hydrocarbon levels for the idle emission inspection
approach can be attributed to the fact that a timing adjustment was in-
cluded in these procedures; this was not the case for the corresponding
state-lane engine parameter inspection procedure.
Emission profiles for different types of emission control devices are
shown in Figures 5-9a, band c for the idle inspection and adjustment procedure.
The resulting trends in the emission profiles are similar to those for the engine
parameter inspection procedure.
Table 5-4 presents a summary of the emission inspection procedures.
The combined idle modes and loaded HC mode inspection is extreme1v cost-
effective and offers substantial emission reductions, 11% and 15% for HC
and CO, respectively. The good figure of merit results from the high
87
-------
5
~
""--
<'
l 4
VI
.....I
LJ.J
> 3
UJ
.....I
Z
0 2
VI
VI
~
UJ
U
I
0
HC
-- --- ...
--
~- -
BASE LINE FLEET
- - - COMPOSITE TEST FLEET
PANEL A
~ 40
""--
<'
l
VI
.....I
UJ
>
UJ
.....I 20
Z
o
PANEL B
CO
--
VI
VI
BASE LINE FLEET
--- COMPOSITE TEST FLEET
10
~
UJ
o
u
o
~ 8
""--
<'
l 6
VI
.....I
UJ
>
UJ 4
.....I
Z
0
VI 2
VI
~
UJ
0 0
Z
NO
~-
--- t----I-
BASE LINE FLEET
- -- COMPOSITE TEST FLEET
PANEL C
2 3
TIME - YEARS
4
5
Fi gure 5-8.
Emission Time Histories for an Idle Emission
and Adjustment Program
88
-------
V1
W
-I
~
""'-
V1
~ 3
~
~
l
V1
-I
OJ W
>
1.0 W
-I 2
Z
0
V1
V1
~
W
EMISSION SIGNATURE ANALYSIS
IDLE ADJUSTMENT PROGRAM
PAN E L A
5
HC - EMISSIONS
.--.-:0= ._._.~..:=
---
- ---- -----------
--- ...::;;=- . -== ~.---
--
--~-=-~-
~~.
~ ,..... .
~.
---
BASE LINE
------ AIR
. EM
- - AVERAGE
4
o
o
3
2
TIME ~ YEARS
4
1
Fi gure 5-9a.
Power Train Emission Histories - Panel A
-------
- - - - .-......:r . i-
..... --- .~,......... ..r..:.; .:.:.:. . . .. . ...7":':':-. -.
.,.".""'.........: -""'......... ........... ....:.::---.-.-.
'....... . ......... ..............
30 .."..,.-....~. ----
.~. ....................
50
40
IJ.J
-I
~
"'-.
V1
~
«
~
0
~
V1
-I
'" IJ.J
>
0 IJ.J
-I 20
Z
o
V1
V1
~
IJ.J
EMISSION SIGNATURE ANALYSIS
IDLE MAINTENANCE PROGRAM
PAN E L B
CO EMISSIONS
BASE LINE
---- - AIR
-.-.. EM
............ AVERAGE
10
o
o
2
TIME - YEARS
3
1
Fi gure 5-9b.
Power Train Emission Histories - Panel B
4
-------
8
w
-J
~
~
V'J
~
<{
eo=:
(9
l
V'J
1.0 -J
-...I W
>
W
-J
Z
o
V'J
V'J
~
W
PANEL C
EMISSION SIGNATURE ANALYSIS
IDLE MAINTENANCE PROGRAM
10
NO EMISSIONS I
BASE LINE
-- - -- AIR
-.-. EM
... .. ... . .. AVERAGE
1-.-
.-.-.
-.-.-
.-.- ~._._._.- r---.-.-.-
............. .................... ...
........................ ...... .. .... . .........
~---------- ----------
~-------- --------
6
4
2
o
o
1
2
TIME - YEARS
3
4
Fi gure 5-9c.
Power Train Emission Histories - Panel C
-------
weighting factor (0.6) on HC and significantly lower time regime for
inspection (one additional minute) when compared to the conventional
scope diagnosis which takes 15 minutes. The lower emission reductions
when induction system repairs are included in the procedures are due to
the large number of errors of omission. These omission errors, in
general, tend to reduce the effectiveness of the several inspection
procedures. Errors of commission directly impact on the costs associated
with operating the program. These higher program costs result from the
added time required to reinspect each vehicle that failed the initial
screening inspection.
92
-------
Table 5-4.
Optimal Emission Signature Inspection Strategies
4-Year Average Emission Reductions
Weight Factors: HC = 0.6, CO = 0.1, NO = 0.3
~
w
I N S P EC TI ON % EM I SS I ON RE DUCT I ON
MO DE S UBS Y STEM F I GU RE .0 F ME RI T CO ST PE R I NS PEC TI ON HC CO NO
$/TO N $
n-
I I DL E CO I DL E PARAME TE RS 4 22 2 . 50 2 1 2 - 4
I I DL E ( CO + HC ) I DL E PARAMETE RS 458 2 . 50 3 1 1 -4
39 4 2 . a a -2 1 2 -4
I I DL E ( CO + H C ) I DL E PA RAM ET E RS
+ L OA DE D ( HC ) I GN I T I ON tttttt~i~:ttt:~ttttttt\:ttt::t:tt:tt::::::::::::::::::::::::::::::::::::::::::::!:~:mE:t:t:t:t:t:tI:::::::t:t:t:t:t:t:t:::t:::::tt:t:::t:::::::j:'t!t:t:t:t:t::,:::::t:::t:t:tttttt:jtit:::::t::t::t:::::::::::::::::::::::t~::::
280 * 3 00 1 a 1 3 3
. -
I I DL E ( CO + HC ) I DL E PARAMETE RS
+ La ADE D ( H C+ CO ) I GN I TI a N + ::::::~'~::~:::!t:tt:::~~::::::::ttt:::::::::::::::t::,:!:.:,:::::t:i:':::::::::::::~:~:ijij[[[)::j:::::::::::::::::::::::::::::'1::::::':':':::':,:::::::':::::'::::;~:::
I N DUCT I ON ::::::~:t~:t::ili::::::::::::::::::~:::::::::::::::::t::::::::::::::::::tf:,qmt:::::::::t::::::ffff::::::::::::::::::::::::::::f:~:g:::::::::::::fffIJ:'t:::::::::::::::::::::::t::::f:f~:::
t Co ns t ra i n e d 0 p t i mum to ob ta i n a 1 5% redu c ti on i n H C 0 r CO
* U n c on s t ra i ne d 0 P ti mum
-------
REFERENCES
1.
Vehicle Emissions Surveillance Study, APRAC Project Number CAPE-13-68,
sponsored by the Coordinating Research Council, Inc. and NAPCA,
October 1970.
2.
E. L. Cline and L. Tinkham, "A Realistic Vehicle Emission Inspection
System," APCA Paper No. 68-152.
Chilton's Labor Guide and Parts Manual, Motor/Age, 40th Edition, 1969.
3.
4.
TRW Report No. 09793-6002-ROOO, "Automated Diagnostic Systems-Vehicle
Inspection," Final Report-Phase I on Contract FH-11-6538 for the
National Highway Safety Bureau.
5.
G. W. Dickinson, Ildvad, H. M., R. J. Bergin, Tune-up Inspection a
Continuing Emission Control, Proving Ground Section, General Motors
Corp. SAE Paper No. 690141, January 13-17, 1969.
6.
E. L. Cline, "State of Wisconsin Governor's Workshop on Motor Vehicle
Air Pollution," talk presented to Clayton Manufacturing Co.,
March 25, 1970.
7.
Automobile Club of Southern California, Engineering & Technical
Services Division, Automotive Fleet Emissions Program, June 15, 1968.
8.
F- Mosteller, R. C. Rourke, G. Thomas, Probability with Statistical
Applications, Addison-Wesley Publishing Co., Mass., 1961, p. 81.
Federal Register, ll, no. 61, Part II -"Control of Air Pollution from
New Motor Vehicles and New Motor Vehicle Engines;'March 30, 1966.
9.
10.
Health, Education, and Welfare, Division of Motor Vehicles' Pollution
Control, 1968 and 1969 Surveillance Data.
11.
A.J. Hocker,"Exhaust Emissions from Privately Owned 1966-69 California
Automobiles," CARB Supplement to Progress Report No. 18, March 20, 1970.
94
------- |